EA046398B1 - POLYSPECIFIC ANTIGEN-BINDING MOLECULES AND THEIR APPLICATIONS - Google Patents

Description

Field of invention

The present invention relates to the field of therapeutic proteins and, in particular, to the field of therapeutic proteins that are capable of inactivating, blocking, weakening, eliminating and/or reducing the concentration of one or more target molecules in vitro or in vivo.

BACKGROUND OF THE INVENTION

Therapeutic modalities often require the inactivation or blocking of one or more target molecules that act near or on the cell surface. For example, antibody therapeutics often function by binding to a specific antigen expressed on the cell surface or to a soluble ligand, thereby interfering with the normal biological activity of the antigen. Antibodies and other binding constructs directed against various cytokines (eg, IL-1, IL-4, IL-6, IL-13, IL-22, IL-25, IL-33, etc.) or their respective receptors , for example, have been shown to be useful in the treatment of a wide range of human disease states and diseases. Therapeutics of this type typically function by blocking the interaction between a cytokine and its receptor to attenuate or inhibit cellular signal transduction. However, in certain cases, it would be therapeutically advantageous to inactivate or inhibit the activity of a target molecule in a manner that does not necessarily involve blocking its physical interaction with another moiety. One way to achieve such attenuation that does not involve blocking the target molecule would be to reduce the concentration of the target molecule outside the cell or on the cell surface. Although genetic and nucleic acid-based strategies for reducing the amount or concentration of a given target molecule are known in the art, such strategies are often fraught with significant technical difficulties and undesirable side effects in therapeutic settings. Accordingly, alternative non-blocking strategies are needed to promote inactivation or attenuation of various target molecules for therapeutic purposes.

Short description

The present invention is based, at least in part, on the concept of attenuating or inactivating a target molecule by facilitating or providing physical association between the target molecule and an internalizing effector protein. Through this type of physical intermolecular communication, the target molecule can be forced into the cell along with the internalizing effector protein and processed through an intracellular degradation mechanism or otherwise attenuated, sequestered, or inactivated. This mechanism represents a new and proposed strategy for inactivating a target molecule or reducing its activity without the need to block the interaction between the target molecule and its binding partners.

Accordingly, the present invention provides a multispecific antigen binding molecule that is capable of simultaneously binding a target molecule (T) and an internalizing effector protein (E). More specifically, the present invention provides a multispecific antigen binding molecule comprising a first antigen binding domain (D1) and a second antigen binding domain (D2), wherein D1 specifically binds T and D2 specifically binds E, and wherein simultaneous binding of T and E by the polyspecific antigen binding molecule attenuates the activity of T to a greater extent than T binding through D1 alone. Increased attenuation of T activity may be due to forced internalization/degradation of T due to its physical connection with E; however, other mechanisms of action are possible and are not excluded from the scope of the present invention.

In addition, the present invention provides methods for using a multispecific antigen binding molecule to inactivate or reduce the activity of a target molecule (T). In particular, the present invention provides a method of inactivating or attenuating T activity by bringing T and an internalizing effector protein (E) into contact with a polyspecific antigen binding molecule, wherein the polyspecific antigen binding molecule comprises a first antigen binding domain (D1) and a second antigen binding domain (D2), wherein D1 specifically binds T and where D2 specifically binds E; and where simultaneous binding of T and E by a multispecific antigen binding molecule attenuates T activity to a greater extent than binding of T by D1 alone.

In certain embodiments of the present invention, D1 and/or D2 comprise(s) at least one antibody variable region. For example, a multispecific antigen binding molecule may, in some embodiments, be a bispecific antibody, wherein D1 contains a pair of antibody heavy and light chain variable regions (HCVR/LCVR) that specifically binds T, and wherein D2 contains a pair HCVR/LCVR that specifically binds E Alternatively, D1 and/or D2 may contain a peptide or polypeptide that specifically interacts with a target molecule (T) and/or an internalizing effector protein (E). For example, if the target molecule is a cell surface receptor, then D1

- 1 046398 may contain part of a ligand that specifically binds the target molecule, which is a cell surface receptor. Likewise, if the internalizing effector protein is a cell surface internalizing receptor, then D2 may contain a portion of a ligand that specifically binds the internalizing cell surface receptor. In certain embodiments, D1 contains an antibody variable region that specifically binds T, and D2 contains a peptide or polypeptide that specifically binds E. In yet other embodiments, D1 contains a peptide or polypeptide that specifically binds T, and D2 contains an antibody variable region, which specifically binds E. However, in any configuration, the end result is that T and E are able to physically bind, either directly or indirectly, due to the simultaneous binding of T and E by a multispecific antigen-binding molecule.

Other embodiments will be apparent from a review of the following detailed description.

Brief description of graphic materials

In fig. 1 (Panels A-D) are schematic representations of four general exemplary mechanisms of action for the multispecific antigen binding molecules of the present invention. In each configuration illustrated, D1 represents the first antigen binding domain; D2 is the second antigen binding domain; T represents the target molecule; E is an internalizing effector protein; and R is a receptor that is internalized upon binding of E. Panel A depicts a situation in which T and E are both membrane bound. Panel B depicts a situation in which T is soluble and E is membrane bound. Panel C depicts a situation in which T is membrane bound and E is a soluble protein that interacts and is internalized into the cell through the interaction of E and R. Panel D depicts a situation in which T is soluble and E is a soluble protein that interacts and internalized into the cell through the interaction of E and R.

In fig. Figure 2 shows the results of an immunoprecipitation experiment performed on two different cells (Cell-1, expressing only FcyRI, and Cell-2, expressing Krm2 and FcyRI), after incubation for different periods of time (0, 15, 30 and 60 min) with polyspecific antigen-binding molecule DKK1-mFc.

In fig. Figure 3 shows the relative IL-4-induced luminescence produced by HEK293 cells with the Stat6-luc reporter system in the presence and absence of polyspecific antigen binding protein binding to IL-4R/CD63 (ab-conjugate) or control constructs (control 1 and control 2). at different concentrations of IL-4.

In fig. 4 shows the results of an experiment carried out in the same manner as the experiment shown in FIG. 3, except that CD63 expression was significantly reduced in the reporter cell line by siRNA directed against CD63.

In fig. 5 shows the results of an experiment carried out in the same manner as the experiments shown in FIG. 3 and 4, except that reporter cells were incubated with polyspecific antigen binding protein (Ab conjugate) or control constructs (control 1 and control 2) at time zero (panels A and B), for 2 hours (panels C and D) or overnight (panels E and F) before adding IL-4 ligand. The top row of histograms (panels A, C, and E) represents the results of experiments performed with cells expressing normal levels of CD63 (untransfected), whereas the bottom row of histograms (panels B, D, and F) represents the results of experiments performed with cells in which CD63 expression was significantly reduced in the reporter cell line by siRNA directed against CD63.

In fig. 6 shows the results of an experiment carried out in the same manner as the experiments shown in FIG. 3 and 4, except that reporter cells were incubated with IL-4R/CD63 polyspecific antigen binding protein (Ab conjugate) or control constructs (control 1 and control 2) for 15 min (panel A), 30 min (panel B), 1 h (panel C), or 2 h (panel D) before adding IL-4 ligand.

In fig. Figure 7 shows the results of an experiment in which cells with the Stat6-luc reporter system were treated with 10 pM IL-4 in the presence of various dilutions of the bispecific antibody to IL-4RxCD63 (bispecific) or control constructs (IL-4R monospecific or mock bispecific that binds only IL -4R).

In fig. 8 shows the results of experiments in which HEK293 cells were treated with a tyc-tagged and pH-sensitive SOST construct (which produces a fluorescent signal at low pH), as well as various monospecific and bispecific antibodies, such as those shown. Results are expressed as the number of fluorescent spots (i.e., labeled vesicles) per cell. Panel A shows the results obtained after incubation on ice for 3 hours, panel B shows the results after 1 hour of incubation at 37°C, and panel C shows the results after 3 hours of incubation at 37°C.

In fig. Figure 9 shows the results of experiments in which HEK2 93 cells were treated with fluorescently labeled lipopolysaccharide (LPS) from E. coli (panel A) or S. minnesota (panel B), along with

- 2 046398 with a bispecific antibody to CD63xLPS, control antibodies or LPS alone for various periods of time, followed by quenching of the non-internalized (ie surface bound) fluorophore. Thus, the fluorescent signal reflects internalized LPS under the various conditions shown. Results are expressed as the number of fluorescent spots (i.e., labeled vesicles) per cell.

In fig. Figure 10 shows the average fluorescence values in arbitrary units from Alexa488-labeled FelD1-mycv-myc-his. Blue histograms show cell surface labeling. The red histograms depict the internalized label. Group 1 on the x-axis represents cells treated with bispecific antibody to HLA-BxFelD1; group 2 represents cells treated with parental bivalent monospecific HLA-B antibody; group 3 represents treatment with IgG isotype controls. Panel A shows the binding and internalization of FelD1-mmh-488 by Blymphoblastoid CIRneo cells that do not express MHC1. Panel B shows the binding and internalization of FelD1-mmh-488 by CIRneo B lymphoblastoid cells that express MHC1.

In fig. 11 shows the internalization of prolactin receptor (PRLR) and HER2 on T47D cells by measuring the amount of PRLR or HER2 remaining on the cell surface 0-60 minutes after cells were transferred from 4 to 37°C. The percentage of surface receptor remaining over time is shown. Squares represent PRLR and triangles represent HER2.

In fig. 12 depicts co-localization of prolactin receptor (PRLR) and HER2 in lysosomes in T47D cells. The percentage of receptor bound to lysosomes over time is shown. Squares represent PRLR and triangles represent HER2.

In fig. 13 schematically depicts PRLR and HER2 truncations and chimeras.

In fig. 14 shows fluorescence micrographs of HEK293 cells expressing various PRLR and HER2 constructs and chimeras. The subpanels on the left of each panel show cells at 4°C before internalization. The subpanel on the right of each panel shows cells after 1 h at 37°C. Panel A shows cells expressing full-length PRLR (PRLR FL). Panel B shows cells expressing full-length HER2 (PRLR FL). Panel C shows cells expressing the PRLRectoHER2cyto™ construct. Panel D shows cells expressing the HER2ectoPRLRcyto™ construct.

In fig. 15 shows fluorescence micrographs of HEK293 cells expressing various PRLR constructs and truncations. Panel A shows cells expressing full-length PRLR stained for PRLR after 1 h at 4°C. Panel B shows cells expressing full-length PRLR stained for PRLR after 1 h at 37°C. Panel C shows cells expressing a truncated PRLR with 42 cytoplasmic domain residues stained for PRLR after 1 h at 4°C. Panel D shows cells expressing a truncated PRLR with 42 cytoplasmic domain residues stained for PRLR after 1 h at 37°C. Panel E shows cells expressing a truncated PRLR with only the 21-residue cytoplasmic domain stained for PRLR after 1 h at 4°C. Panel F shows cells expressing a truncated PRLR with only the 21-residue cytoplasmic domain stained for PRLR after 1 h at 37°C.

In fig. 16 shows Western blots of cell lysates of cells expressing full-length PRLR (panel A), cells expressing truncated PRLR with 42 residues of the remaining cytoplasmic domain (panel B), and cells expressing truncated PRLR with 21 residues of the remaining cytoplasmic domain (panel C). The upper subpanels show anti-PRLR antibody staining. The lower subpanels show beta-actin staining to control loading.

In fig. 17 shows the percentage of cells in early mitosis compared to the expression levels of PRLR or HER2 or PRLRectoHER2cyto™ (panel A), PRLR or HER2ectoPRLRcyto™ (panel B) on the cell surface after treatment with either PRLR-DM1 or HER2-DM1.

In fig. 18 shows a Western blot of total cell lysates expressing various truncations and substitutions of PRLR constructs at 0 and 4 hours after CHX treatment. The top panel shows PRLR staining and the bottom panel shows beta-actin staining for loading control.

In fig. 19 shows a Western blot of total cell lysates from HEK 293 cells induced to express full-length PRLR at 0, 1, 2 and 4 hours after CHX treatment. The top panel shows HER2 staining and the bottom panel shows PRLR staining.

In fig. 20 shows a Western blot of total cell lysates from HEK 293 cells induced to express the cytoplasmic truncation form of PRLR at 0, 2 and 4 hours after CHX treatment. The top panel shows HER2 staining and the bottom panel shows PRLR staining.

In fig. Figure 21 shows co-localization of HER2 with lysosomes in T47D cells treated with a bispecific antibody to HER2xPRLR. The percentage of lysosomes bound to the HER2 receptor is shown over time. Squares represent cells treated with bispecific antibody to HER2xPRLR, and triangles represent cells treated with nonbinding control antibodies.

- 3 046398

Fig. 22 is a bar graph showing the percentage of T47D/HER2 cells with cell cycle arrest (Y-axis) treated with 1, 10, or 30 nM PRLR-ADC (A), HER2-ADC plus anti-HER2xPRLR bispecific antibody (B), HER2- only ADC (C), nonbinding control ADC (D), or no treatment (E).

In fig. 23 is a dot plot showing the percentage of viable T47D cells expressing HER2 as a function of increasing drug amounts: PRLR-ADC (A; squares), HER2-ADC plus anti-HER2xPRLR bispecific antibody (B; diamonds), HER2-ADC only (C; point-up triangles), ADC non-binding control (D; circles), or ADC non-binding control plus HER2xPRLR bispecific antibody (E; point-down triangles).

In fig. 24 shows a panel of Western blot results of mouse serum probed with anti-FelD1 antibody at various time points (15 minutes, 6 hours, 1, 2, 3, 4 and 6 days) after treatment with bispecific anti-HLABxFelD1 antibody, phosphate buffered saline, bivalent monospecific anti-FelD1 antibody and bivalent monospecific anti-HLAB antibody.

In fig. 25 is a histogram depicting the proportion of FelD1-Fc recovered from mouse serum above baseline levels and normalized to baseline (Y-axis). Number 1 shows treatment with a bispecific antibody to HLABxFelD1; number 2 shows treatment with bivalent monospecific antibody to FelD1 and number 3 shows treatment with bivalent monospecific antibody to HLAB (X-axis).

Fig. 26 is a bar graph showing the uptake of pHrodo®-hHJV-mmh by HEK293 cells. The Y-axis shows the integrated intensity (in arbitrary units) of the pHrodo® signal. On the X axis, Anti-Myc::PCSK9 shows the effect on HJV uptake of the absence of antibody (open bar), a-Myc::PCSK9FL (fully filled bar), α-Myc::PCSK9LC (horizontal filled bar), and α-Myc::PCSK9LC (horizontal filled bar), and α-Myc::PCSK9LC (horizontal filled bar), and α-Myc::PCSK9LC (horizontal filled bar), and α-Myc::PCSK9LC: :PCSK9SC (shaded bar); Anti-HJV::PCSK9 demonstrates the effect on HJV uptake by the absence of antibody (open bar), α-HJV-N::PCSK9FL (fully filled bar), a-HJV-N::PCSK9LC (horizontal filled bar), and a-HJV -N::PCSK9SC (shaded bar) and controls without antibody for background absorbance.

Fig. 27 is a dot plot showing serum iron levels in micrograms per deciliter of serum (Y-axis) one week after treatment of hHLAB transgenic mice with an anti-HJV blocking bivalent monospecific antibody (1); full-length anti-Myc::PCSK9 fusion protein (2); a non-blocking anti-HJV antibody bivalent monospecific antibody (3) and a non-blocking anti-HJV::PCSK9 full-length fusion protein (4).

Detailed Description of the Invention

Before describing the present invention, it should be understood that the present invention is not limited to the specific methods and experimental conditions described, since such methods and conditions may vary. It should also be understood that the terminology used herein is for purposes of describing specific embodiments only and is not intended to be limiting as the scope of the present invention will be limited only by the appended claims.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one skilled in the art to which the present invention belongs. As used herein, the expression approximate, when used in relation to a specific specified numerical value, means that the value may differ from the specified value by no more than 1%. For example, as used herein, the expression approximately 100 includes 99 and 101 and all values in between (eg, 99.1, 99.2, 99.3, 99.4, etc.).

Although any methods and materials similar or equivalent to those described herein may be used in practicing or testing the present invention, only the preferred methods and materials are described herein. All patents, applications and non-patent publications mentioned in this specification are incorporated herein by reference in their entirety.

Polyspecific antigen-binding molecules.

The present inventors have surprisingly discovered that the activity of a target molecule can be attenuated by binding the target molecule to an internalizing effector protein via a multispecific antigen binding molecule.

Accordingly, the present invention provides multispecific antigen binding molecules comprising a first antigen binding domain (also referred to herein as D1) and a second antigen binding domain (also referred to herein as D2). D1 and D2 bind different molecules. D1 specifically binds the target molecule. The target molecule is also

- 4 046398 is referred to herein as T. D2 specifically binds internalizing effector protein.

The internalizing effector protein is also referred to herein as E. In accordance with the present invention, simultaneous binding of T and E by a multispecific antigen binding molecule attenuates the activity of T to a greater extent than binding of T by D1 alone. As used herein, the expression simultaneous binding in the context of a polyspecific antigen binding molecule means that the polyspecific antigen binding molecule is capable of contacting both the target molecule (T) and the internalizing effector protein (E) for at least some period of time in a physiologically appropriate manner. conditions to promote physical association between T and E. Binding of a multispecific antigen-binding molecule to components T and E may be sequential; for example, a polyspecific antigen binding molecule may bind T first and then bind E, or may bind E first and then bind T. In either case, as long as T and E are both bound by the polyspecific antigen binding molecule for some period of time (regardless of the binding sequence) , a multispecific antigen binding molecule will be considered to simultaneously bind T and E for the purposes of the present invention. Without being bound by theory, it is believed that enhanced T inactivation is caused by internalization and redirection to degradation of T in the cell due to its physical association with E. Thus, the multispecific antigen binding molecules of the present invention are useful for inactivating and/or reducing activity and/or concentration outside the cell target molecule without directly blocking or disrupting the function of the target molecule.

In accordance with the present invention, the polyspecific antigen binding molecule may be a single multifunctional polypeptide or may be a multimeric complex of two or more polypeptides covalently or non-covalently linked to each other. As will be clear from the present invention, any antigen binding construct that has the ability to simultaneously bind T and E molecules is considered a multispecific antigen binding molecule. Any of the multispecific antigen binding molecules of the present invention or variants thereof can be constructed using standard molecular biology techniques (eg, recombinant DNA technology and protein expression) known to one of ordinary skill in the art.

Antigen binding domains.

The polyspecific antigen binding molecules of the present invention contain at least two distinct antigen binding domains (D1 and D2). As used herein, the expression antigen-binding domain means any peptide, polypeptide, nucleic acid molecule, framework molecule, peptide display molecule, or polypeptide-containing construct that is capable of specifically binding to a particular antigen of interest. The term specifically binds or the like as used herein means that the antigen binding domain forms a complex with a particular antigen having a dissociation constant (KD) of 500 pM or less, and does not bind other unrelated antigens under normal testing conditions. Unrelated antigens are proteins, peptides or polypeptides that have less than 95% amino acid identity to each other.

Exemplary categories of antigen binding domains that may be used in the context of the present invention include antibodies, antigen binding portions of antibodies, peptides that specifically interact with a particular antigen (e.g., peptibodies), receptor molecules that specifically interact with a particular antigen, proteins containing a ligand binding portion of a receptor , which specifically binds a specific antigen, antigen-binding scaffolds (e.g., DARPin, HEAT repeat proteins, ARM repeat proteins, tetratricopeptide repeat proteins, and other naturally occurring repeat protein scaffolds, etc. (see, e.g., Boersma and Pluckthun, 2011, Curr. Opin. Biotechnol., 22:849-857 and references therein)) and aptamers or parts thereof.

In certain embodiments in which the target molecule or internalizing effector protein is a receptor molecule, the antigen binding domain for purposes of the present invention may contain or consist of a ligand or part of a ligand that is specific for the receptor. For example, if the target molecule (T) is IL-4R, the D1 component of the polyspecific antigen binding molecule may contain an IL-4 ligand or a portion of an IL-4 ligand that is capable of specifically interacting with IL-4R; or if the internalizing effector protein (E) is a transferrin receptor, the D2 component of the multispecific antigen binding molecule may contain transferrin or a portion of transferrin that is capable of specifically interacting with the transferrin receptor.

In certain embodiments in which the target molecule or internalizing effector protein is a ligand that is specifically recognized by a particular receptor (eg, a soluble target molecule), the antigen binding domain for purposes of the present invention may comprise or consist of a receptor or a ligand binding portion of a receptor. For example,

- 5 046398 if the target molecule (T) is IL-6, the D1 component of the polyspecific antigen binding molecule may contain an IL-6 receptor ligand binding domain; or if the internalizing effector protein (E) is an internalized indirect protein (as that term is defined elsewhere herein), the D2 component of the polyspecific antigen binding molecule may comprise an E-specific receptor ligand binding domain.

Methods for determining whether two molecules specifically bind to each other are well known in the art and include, for example, equilibrium dialysis, surface plasmon resonance, and the like. For example, an antigen binding domain used in the context of the present invention includes polypeptides that bind a specific antigen (eg, a target molecule [T] or an internalizing effector protein [E]) or a portion thereof with a KD of less than about 500 pM, less than about 400 pM , less than about 300 pM, less than about 200 pM, less than about 100 pM, less than about 90 pM, less than about 80 pM, less than about 70 pM, less than about 60 pM, less than about 50 pM, less less than about 40 pM, less than about 30 pM, less than about 20 pM, less than about 10 pM, less than about 5 pM, less than about 4 pM, less than about 2 pM, less than about 1 pM, less than about 0.5 pM, less than about 0.2 pM, less than about 0.1 pM, or less than about 0.05 pM, as measured by surface plasmon resonance analysis.

The term surface plasmon resonance as used herein refers to an optical phenomenon that provides the ability to analyze interactions in real time by detecting changes in protein concentrations in a biosensor array, for example, using the BIAcore™ system (Biacore Life Sciences division of GE Healthcare, Piscataway, New Jersey).

As used herein, the term KD refers to the equilibrium dissociation constant of a particular protein-protein interaction (eg, antibody-antigen interaction). Unless otherwise stated, _KD values disclosed herein refer to _KD values determined by surface plasmon resonance at 25°C.

Antibodies and antigen-binding fragments of antibodies.

As stated above, the antigen binding domain (D1 and/or D2) may contain or consist of an antibody or an antigen binding fragment of an antibody. As used herein, the term antibody means any antigen-binding molecule or molecular complex containing at least one complementarity determining region (CDR) that specifically binds or interacts with a particular antigen (eg, T or E). The term antibody includes immunoglobulin molecules containing four polypeptide chains, two heavy (H) chains and two light (L) chains linked by disulfide bonds, as well as their multimers (for example, IgM). Each heavy chain contains a heavy chain variable region (abbreviated herein as HCVR or _VH ) and a heavy chain constant region. The heavy chain constant region contains three domains, _CH1 , CH2 and CH3. Each light chain contains a light chain variable region (abbreviated herein as LCVR or V _L ) and a light chain constant region. The light chain constant region contains one domain (C _L 1). The _VH and VL regions can be further subdivided into regions of hypervariability called complementarity determining regions (CDRs), interspersed with more conserved regions called framework regions (FRs). Each VH and V _L consists of three CDRs and four FRs, located from the amino terminus to the carboxy terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. In various embodiments of the present invention, the FRs of the antibodies of the present invention (or the antigen binding portion thereof) may be identical to human germline sequences or may be naturally or artificially modified. The amino acid consensus sequence can be determined based on side-by-side analysis of two or more CDRs.

The D1 and/or D2 components of the polyspecific antigen binding molecules of the present invention may contain or be composed of antigen binding fragments of full antibody molecules. The terms antigen-binding portion of an antibody, antigen-binding fragment of an antibody, and the like as used herein include any naturally occurring, enzymatically produced, synthetic, or genetically engineered polypeptide or glycoprotein that specifically binds an antigen to form a complex . Antigen-binding antibody fragments can be produced, for example, from whole antibody molecules using any suitable standard methods, such as proteolytic digestion or recombinant genetic engineering technologies involving manipulation and expression of DNA encoding antibody variable and optionally constant domains. Such DNA is known and/or readily available, for example, from commercial sources, DNA libraries (including, for example, phage antibody libraries), or can be synthesized. The DNA can be sequenced and chemically manipulated or manipulated using molecular biology techniques, for example, to arrange one or more variable and/or constant domains into a suitable configuration, or to introduce codons, incorporate cysteine residues, modify, add, or delete

- 6 046398 amino acids, etc.

Non-limiting examples of antigen binding fragments include (i) Fab fragments;

(ii) F(ab')2 fragments;

(iii) Fd fragments;

(iv) Fv fragments;

(v) single chain Fv molecules (scFv);

(vi) dAb fragments; and (vii) minimal recognition units consisting of amino acid residues that mimic the hypervariable region of an antibody (eg, a dedicated complementarity determining region (CDR), such as the CDR3 peptide) or the restricted FR3-CDR3-FR4 peptide.

Other engineered molecules such as domain-specific antibodies, single-domain antibodies, domain-deleted antibodies, chimeric antibodies, CDR-grafted antibodies, diabodies, tri-bodies, tetrabodies, minibodies, nanobodies (e.g. monovalent antibodies, bivalent antibodies, etc.) , small modular protein (SMIP) immunopharmaceuticals and shark IgNAR variable domains are also included in the expression antigen binding fragment used herein.

The antigen binding fragment of an antibody typically contains at least one variable domain. The variable domain can be of any size or amino acid composition and will typically contain at least one CDR that is in frame with or adjacent to one or more framework sequences. In antigen binding fragments containing a VH domain linked to a VL domain, the _VH and VL domains may be arranged relative to each other in any suitable order. For example, the variable region may be dimeric and contain VH-VH, VH-VL or VL-VL dimers. Alternatively, the antigen binding fragment of the antibody may comprise a monomeric VH or VL domain.

In certain embodiments, an antigen binding fragment of an antibody may comprise at least one variable domain covalently linked to at least one constant domain. Non-limiting exemplary configurations of variable and constant domains that may be found in an antigen binding fragment of an antibody of the present invention include (i) Vh-Ch1;

(ii) VH-CH2;

⁽ⁱⁱⁱ⁾ VH-CH3;

(iv) Vh-Ch1-Ch2;

(v) Vh-Ch1-Ch2-Ch3;

(vi) Vh-Ch2-Ch3;

(vii) Vh-Cl;

(viii) Vl-Ch1;

(ix) Vl-Ch2;

(x) Vl-Ch3;

(xi) Vl-Ch1-Ch2;

(xii) Vl-Ch1-Ch2-Ch3;

(xiii) Vl-Ch2-Ch3; and (xiv) Vl-Cl.

In any configuration of variable and constant domains, including any of the exemplary configurations set forth above, the variable and constant domains may either be directly linked to each other or may be linked by all or part of a hinge or linker region. The hinge region may consist of at least 2 (eg, 5, 10, 15, 20, 40, 60 or more) amino acids that provide flexible or semiflexible linkage between adjacent variable and/or constant domains in a single polypeptide molecule. In addition, the antigen binding fragment may comprise a homodimer or heterodimer (or other multimer) of any of the variable and constant domain configurations set forth above, non-covalently linked to each other and/or to one or more monomeric _VH or VL domains (e.g. using disulfide bond(s).

The polyspecific antigen binding molecules of the present invention may contain or consist of human antibodies and/or recombinant human antibodies or fragments thereof. The term human antibody as used herein includes antibodies containing variable and constant regions derived from human germline immunoglobulin sequences. The human antibodies of the present invention may nevertheless include amino acid residues not encoded by human germline immunoglobulin sequences (e.g., through mutations introduced by random or site-specific mutagenesis in vitro or introduction of somatic mutation in vivo), for example in the CDR and , in particular in CDR3. However

- 7046398 The term human antibody as used herein is not intended to include antibodies in which CDR sequences derived from the germline of another mammalian species, such as a mouse, are grafted onto human framework sequences.

The polyspecific antigen binding molecules of the present invention may contain or consist of recombinant human antibodies or antigen binding fragments thereof.

As used herein, the term recombinant human antibody is intended to include all human antibodies produced, expressed, generated or released by recombinant means, such as antibodies expressed using a recombinant expression vector transfected into a host cell (further described below), antibodies, isolated from a recombinant combinatorial human antibody library (further described below), antibodies isolated from an animal (eg, mouse) that is transgenic for human immunoglobulin genes (see, for example, Taylor et al. (1992), Nucl. Acids Res. , 20:6287-6295), or antibodies produced, expressed, created or secreted by any other method that involves joining human immunoglobulin gene sequences with other DNA sequences. Such recombinant human antibodies contain variable and constant regions derived from human germline immunoglobulin sequences. However, in certain embodiments, such recombinant human antibodies are subject to in vitro mutagenesis (or, if a transgenic animal is used for human Ig sequences, to in vivo somatic mutagenesis) and, therefore, the amino acid sequences of the VH and VL regions of the recombinant antibodies are sequences that, being derived from or related to human germline VH and VL sequences may not exist naturally in the human germline antibody repertoire in vivo.

Bispecific antibodies.

In accordance with certain embodiments, the polyspecific antigen binding molecules of the present invention are bispecific antibodies; for example, bispecific antibodies containing an antigen-binding arm that specifically binds a target molecule (T) and an antigen-binding arm that specifically binds an internalizing effector protein (E). Methods for producing bispecific antibodies are known in the art and can be used to construct the multispecific antigen binding molecules of the present invention.

Exemplary bispecific formats that may be used in the context of the present invention include, but are not limited to, for example, scFv-based bispecific or diabody formats, IgG-scFv fusions, double variable domain (DVD)-Ig, quadroma, peak-to-valley, conventional light chain (e.g. regular peak-to-trough light chain, etc.), CrossMab, CrossFab, (SEED)-body, leucine zipper, DuoBody, IgG1/IgG2, Fab (DAF)-Double-acting IgG and bispecific Mab ² formats (see, for example, Klein et al., 2012, mAbs, 4:6, 1-11, and sources cited herein for an overview of the above formats).

Multimerization components.

The polyspecific antigen binding molecules of the present invention, in certain embodiments, may also contain one or more multimerization components. Multimerization components may function to maintain communication between the antigen-binding domains (D1 and D2). As used herein, a multimerization component is any macromolecule, protein, polypeptide, peptide, or amino acid that has the ability to bind to a second multimerization component of the same or similar structure or composition. For example, the multimerization component may be a polypeptide containing an immunoglobulin Cip domain.

A non-limiting example of a multimerization component is the Fc portion of an immunoglobulin, for example, the Fc domain of an IgG selected from the isotypes IgG1, IgG2, IgG3, and IgG4, as well as any allotype within each isotype group. In certain embodiments, the multimerization component is an Fc fragment or amino acid sequence ranging from 1 to about 200 amino acids in length containing at least one cysteine residue. In other embodiments, the multimerization component is a cysteine residue or a short cysteine-containing peptide. Other multimerization domains include peptides or polypeptides containing or consisting of a leucine zipper, a helix-loop motif, or a supercoil motif.

In certain embodiments, the polyspecific antigen binding molecules of the present invention comprise two multimerization domains, M1 and M2, wherein D1 is attached to M1 and D2 is attached to M2, and where the association of M1 with M2 promotes the physical association of D1 and D2 with each other in one polyspecific antigen binding molecule. In certain embodiments, M1 and M2 are identical to each other. For example, M1 may be an Fc domain with a specific amino acid sequence, and M2 is an Fc domain with the same amino acid sequence as M1. Alternatively, M1 and M2 may differ from each other at one or more amino acid positions. For example, M1 may contain the first immunoglobulin (Ig) C _n 3 domain, and M2

- 8 046398 may contain a second Ig _{C H} 3 domain, wherein the first and second Ig _{C H} 3 domains differ from each other in at least one amino acid, and where at least one amino acid difference reduces the ability of the targeted construct to bind to protein A compared to a reference construct having identical M1 and M2 sequences. In one embodiment, the Ig C.'113 domain in M1 binds Protein A, and the Ig C.'113 domain in M2 contains a mutation that reduces or eliminates Protein A binding, such as the H95R modification (IMGT exon numbering; EU numbering H435R ). CH3 M2 may additionally contain modification Y96F (according to IMGT; Y436F according to EU). Additional modifications that can be found within CH3 M2 include D16E, L18M, N44S, K52N, V57M and V82I (by IMGT; D356E, L358M, N384S, K392N, V397M and V422I by EU) in the case of the IgG1 Fc domain; N44S, K52N and V82I (IMGT; N384S, K392N and V422I according to EU) in case of IgG2 Fc domain and Q15R, N44S, K52N, V57M, R69K, E79Q and V82I (according to IMGT; Q355R, N384S, K392N, V397M, R409K, E419Q and V422I according to EU) in the case of the IgG4 Fc domain.

Internalizing effector proteins (E).

In the context of the present invention, the D2 component of the multispecific antigen binding molecule specifically binds the internalizing effector protein (E).

An internalizing effector protein is a protein that is capable of being internalized into a cell, or that otherwise participates in or promotes retrograde membrane transport. In some cases, the internalizing effector protein is a protein that undergoes transcytosis; those. the protein is internalized on one side of the cell and transported to the other side of the cell (e.g., from apical to basal). In many embodiments, the internalizing effector protein is a cell surface expressed protein or a soluble extracellular protein. However, the present invention also contemplates embodiments in which the internalizing effector protein is expressed in an intracellular compartment such as an endosome, endoplasmic reticulum, Golgi apparatus, lysosome, etc. For example, proteins involved in retrograde membrane transport (eg, pathways from early/recycling endosomes to the trans-Golgi network) may serve as internalizing effector proteins in various embodiments of the present invention. In either case, binding of D2 to the internalizing effector protein results in the entire polyspecific antigen-binding molecule and any associated molecules (eg, the target molecule associated with D1) also being internalized into the cell. As explained below, internalizing effector proteins include proteins that are directly internalized into the cell, as well as proteins that are indirectly internalized into the cell.

Internalizing effector proteins that are directly internalized into the cell include membrane-bound molecules with at least one extracellular domain (e.g., transmembrane proteins, GPI-anchored proteins, etc.) that undergo cellular internalization and are preferentially processed along the pathway intracellular degradation and/or recycling. Specific non-limiting examples of internalizing effector proteins that are directly internalized into the cell include, for example, CD63, MHC-I (eg, HLA-B27), Kremen-1, Kremen-2, LRP5, LRP6, LRP8, transferrin receptor, LDL receptor , LDL receptor-related protein 1, ASGR1, ASGR2, amyloid precursor protein-like protein 2 (APLP2), apelin receptor (APLNR), MAL (Myelin And Lymphocyte protein, also known as VIP17), IGF2R , Vacuolar type H ⁺ ATPase, diphtheria toxin receptor, folate receptor, glutamate receptors, glutathione receptors, leptin receptors, scavenger receptors (e.g. SCARA15, SCARB1-3, CD36), etc.

In certain embodiments, the internalizing effector protein is prolactin receptor (PRLR). PRLR has been found not only to be a target for certain therapeutic applications, but also to be an effective internalizing effector protein based on its high rate of internalization and turnover. The potential of PRLR as an internalizing effector protein is, for example, illustrated in WO 2015/026907, which shows, inter alia, that antibodies to PRLR are efficiently internalized by PRLR-expressing cells in vitro.

In embodiments in which E is a directly internalized effector protein, the D2 component of the polyspecific antigen binding molecule may be, for example, an antibody or antigen binding fragment of an antibody that specifically binds E, or a ligand or portion of a ligand that specifically interacts with the effector protein. For example, if E is Kremen-1 or Kremen-2, the D2 component may contain or consist of a Kremen ligand (eg, DKK1) or a Kremen-binding portion thereof. As another example, if E is a receptor molecule such as ASGR1, the D2 component may contain or consist of a receptor-specific ligand (eg, asialoorosomucoid [ASOR] or beta-GalNAc) or a receptor-binding portion thereof.

Internalizing effector proteins that are indirectly internalized into the cell include proteins and polypeptides that are not themselves internalized, but are internalized into

- 9 046398 cell after binding or otherwise associating with a second protein or polypeptide that is directly internalized into the cell. Proteins that are indirectly internalized into the cell include, for example, soluble ligands that are capable of binding to an internalizing receptor molecule expressed on the cell surface. A non-limiting example of a soluble ligand that is (indirectly) internalized into a cell through its interaction with an internalizing receptor molecule expressed on the cell surface is transferrin. In embodiments where E is transferrin (or other indirectly internalized protein), binding of D2 to E and interaction of E with the transferrin receptor (or other internalizing receptor molecule expressed on the cell surface) results in all of the polyspecific antigen binding molecule and any molecules bound to it (eg, the target molecule bound by D1) are internalized into the cell simultaneously with the internalization of E and its binding partner.

In embodiments in which E is an indirectly internalized effector protein, such as a soluble ligand, the D2 component of the polyspecific antigen binding molecule may be, for example, an antibody or antigen binding fragment of an antibody that specifically binds E, or a receptor or portion of a receptor that specifically interacts with soluble effector protein. For example, if E is a cytokine, the D2 component may contain or consist of a corresponding cytokine receptor or ligand binding portion thereof.

Target molecules (T).

In the context of the present invention, the D1 component of the polyspecific antigen binding molecule specifically binds the target molecule (T). A target molecule is any protein, polypeptide, or other macromolecule whose activity or extracellular concentration is desired to be reduced, reduced, or eliminated. In many cases, the target molecule to which D1 binds is a protein or polypeptide (ie, a target protein); however, the present invention also includes embodiments wherein the target molecule (T) is a carbohydrate, glycoprotein, lipid, lipoprotein, lipopolysaccharide, or other non-protein polymer or molecule to which D1 binds. In accordance with the present invention, T may be a target protein expressed on the cell surface or a soluble target protein. Target binding by a multispecific antigen binding molecule can take place in the extracellular environment or on the cell surface. However, in certain embodiments, the multispecific antigen binding molecule binds a target molecule within a cell, for example, in an intracellular component such as the endoplasmic reticulum, Golgi apparatus, endosome, lysosome, etc.

Examples of target molecules expressed on the cell surface include receptors expressed on the cell surface, membrane-associated ligands, ion channels, and any other monomeric or polymeric polypeptide component with an extracellular portion that is attached to or associated with the cell membrane. Non-limiting illustrative target molecules expressed on the cell surface that can be targeted by the multispecific antigen binding molecule of the present invention include, for example, cytokine receptors (e.g., IL-1, IL-4, IL-6, IL-13, IL- 22, IL-25, IL-33, etc.), as well as cell surface targets including other type 1 transmembrane receptors such as PRLR, G protein coupled receptors such as GCGR, ion channels such such as Nav1.7, ASIC1 or ASIC2, non-receptor surface proteins such as MHC-I (eg HLA-B*27), etc.

In embodiments in which T is a target protein expressed on a cell surface, the D1 component of the polyspecific antigen binding molecule may be, for example, an antibody or antigen binding fragment of an antibody that specifically binds T, or a ligand or portion of a ligand that specifically interacts with the target a protein expressed on the cell surface. For example, if T is IL-4R, the D1 component may contain or consist of IL-4 or a receptor binding portion thereof.

Examples of soluble target molecules include cytokines, growth factors and other ligands and signaling proteins. A non-limiting exemplary soluble target protein that may be targeted by the multispecific antigen binding molecule of the present invention includes, for example, IL-1, IL-4, IL-6, IL-13, IL-22, IL-25, IL-33, SOST, DKK1, etc. Soluble target molecules also include, for example, non-human target molecules such as allergens (eg, Fel D1, Betv1, CryJ1), pathogens (eg, Candida albicans, S. Aureus, etc.) and pathogenic molecules ( eg lipopolysaccharide [LPS], lipoteichoic acid [LTA], protein A, toxins, etc.). In embodiments in which T is a soluble target molecule, the D1 component of the polyspecific antigen binding molecule may be, for example, an antibody or antigen binding fragment of an antibody that specifically binds T, or a receptor or portion of a receptor that specifically interacts with the soluble target molecule. For example, if T is IL-4, the D1 component may contain or consist of IL-4R or a ligand binding portion thereof.

- 10 046398

Target molecules also include tumor-associated antigens, as described elsewhere herein.

pH-dependent binding.

The present invention provides multispecific antigen binding molecules comprising a first antigen binding domain (D1) and a second antigen binding domain (D2), wherein one or both antigen binding domains (D1 and/or D2) bind their antigen (T or E) depending on pH. For example, the antigen binding domain (D1 and/or D2) may have a reduced ability to bind its antigen at acidic pH compared to neutral pH. Alternatively, the antigen binding domain (D1 and/or D2) may have an increased ability to bind its antigen at acidic pH compared to neutral pH. Antigen-binding domains with pH-dependent binding characteristics can be obtained, for example, by screening a population of antibodies for decreased (or increased) binding ability to a particular antigen at acidic pH compared to neutral pH. In addition, modifications of the antigen binding domain at the amino acid level can produce antigen binding domains with pH-dependent characteristics. For example, by replacing one or more amino acids of an antigen binding domain (eg, within a CDR) with a histidine residue, one can obtain an antigen binding domain with reduced antigen binding capacity at acidic pH compared to neutral pH.

In certain embodiments, the present invention includes multispecific antigen-binding molecules containing a D1 and/or D2 component that binds its corresponding antigen (T or E) at an acidic pH with a KD that is at least about 3, 5, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100 times or more the KD D1 and/or D2 component to bind to the corresponding antigen at neutral pH. pH-dependent binding can also be expressed as the _t ^1/2 of an antigen-binding domain for its antigen at acidic pH compared to neutral pH. For example, the present invention includes multispecific antigen-binding molecules containing a D1 and/or D2 component that binds its corresponding antigen (T or E) at an acidic pH with a t1/ ₂ that is at least about 5, 6, 7. 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60 or more times less than t1/ ₂ D1 and/or D2 component for binding to the corresponding antigen at neutral pH.

The polyspecific antigen-binding molecules of the present invention, which contain a D1 and/or D2 component with reduced antigen-binding capacity at acidic pH compared to neutral pH, when administered to animal subjects may, in certain embodiments, exhibit slower clearance from the bloodstream compared to comparator molecules , which do not exhibit pH-dependent binding characteristics. In accordance with this aspect of the present invention, there are provided polyspecific antigen binding molecules with a reduced ability to bind antigen to T and/or E at acidic pH compared to neutral pH, which are characterized by at least 2 times slower clearance from the bloodstream compared to comparable antigen binding molecules , which do not have a reduced ability to bind antigen at acidic pH compared to neutral pH. The clearance rate can be expressed in terms of the half-life of the antibody, where slower clearance correlates with a longer half-life.

As used herein, the expression acidic pH means a pH of 6.0 or less. The expression acidic pH includes pH values of approximately 6.0, 5.95, 5.8, 5.75, 5.7, 5.65, 5.6, 5.55, 5.5, 5.45, 5 ,4, 5.35, 5.3, 5.25, 5.2, 5.15, 5.1, 5.05, 5.0 or less. As used herein, the expression neutral pH means a pH of from about 7.0 to about 7.4. The expression neutral pH includes pH values of approximately 7.0, 7.05, 7.1, 7.15, 7.2, 7.25, 7.3, 7.35 and 7.4.

Weakening the activity of the target molecule.

As mentioned elsewhere in this document and as shown in the following demonstration examples, the present inventors have discovered that simultaneous binding of a target molecule (T) and an internalizing effector protein (E) by a polyspecific antigen binding molecule attenuates the activity of T to a greater extent than binding of T only to the first component in the form of the antigen-binding domain (D1) of a polyspecific antigen-binding molecule. As used herein, the expression attenuates T activity more than T binding by D1 alone means that in an assay in which T activity can be measured using cells that express E, the level of T activity measured in the presence of a multispecific antigen binding molecule is at least 10% lower than the level of T activity measured in the presence of a control construct containing D1 alone (ie, without physical association with the second antigen binding domain (D2)). For example, the level of T activity measured in the presence of a multispecific antigen binding molecule may be approximately 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90 , 95 or 100% below the level of T activity measured in the presence of a control construct containing D1 alone.

A non-limiting illustrative analysis format for determining whether a polyspecies

- 11 046398 physical antigen binding molecule attenuates the activity of the target molecule to a greater extent than binding of the target molecule through the D1 domain alone is shown in Demonstrations 1 and 2 herein below. In Example 1, for example, T is interleukin-4 receptor (IL-4R) and E is CD63. The polyspecific antigen binding molecule of Example 1 is an antibody conjugate 2 comprising an anti-IL-4R mAb linked to an anti-CD63 mAb via a streptavidin/biotin linker. Thus, D1 in this illustrative construct is the antigen binding domain (HCVR/LCVR pair) of the anti-IL-4R antibody, and D2 is the antigen binding domain (HCVR/LCVR pair) of the anti-CD63 antibody. For the experiments of Examples 1 and 2, a cell-based assay format was used in which a reporter signal is produced by stimulating IL-4R activity by adding exogenous IL-4 ligand. The level of IL-4-induced reporter activity detected in the presence of a polyspecific antigen-binding molecule was compared with the level of IL-4-induced reporter activity detected in the presence of control constructs containing an antibody to IL-4R, either associated with an irrelevant control immunoglobulin (control 1), or in combination with an anti-CD63 antibody (control 2), but without physical connection with it. Thus, control constructs create a condition in which T is bound through D1 alone (ie, where D1 is not part of a multispecific antigen-binding molecule per se). If the level of activity of the target molecule (represented by the reporter signal) observed in the presence of a polyspecific antigen binding molecule is at least 10% less than the level of activity of the target molecule observed in the presence of a control construct containing a D1 component not physically linked to the D2 component (eg, Control 1 or Control 2), then for the purposes of the present invention it is concluded that simultaneous binding of T and E by a multispecific antigen binding molecule attenuates T activity to a greater extent than binding of T by D1 alone.

Binding of T by D1 alone may, in some embodiments, result in partial attenuation of T activity (as in Example 1, where treatment of reporter cells with anti-IL-4R antibody alone (i.e., controls 1 and 2) provided a small level of attenuation of IL signaling -4 compared to untreated cells). In other embodiments, binding of T by D1 alone will result in undetectable attenuation of T activity; those. the biological activity of T may not be affected by binding of T through D1 alone. In any case, however, simultaneous binding of T and E by the multispecific antigen binding molecule of the present invention will attenuate T activity to a greater extent than binding of T by D1 alone.

Alternative assay formats and variations of assay format(s) set forth herein by way of example will be apparent to those skilled in the art given the nature of the particular target molecule and effector proteins to which any given multispecific antigen binding molecule may be applicable. Any such format can be used in the context of the present invention to determine whether simultaneous binding of T and E by a multispecific antigen binding molecule attenuates T activity to a greater extent than binding of T by D1 alone.

Targeted effect on the tumor.

In another aspect of the present invention, multispecific antigen binding molecules are useful for targeting tumor cells. In accordance with this aspect of the present invention, the target T molecule to which D1 binds is a tumor-associated antigen. In certain cases, a tumor-associated antigen is an antigen that is not normally internalized.

The internalizing effector protein E to which D2 binds may be tumor specific or may be expressed on both human tumor and non-tumor cells. Any of the internalizing effector proteins mentioned elsewhere herein may be targeted for antitumor applications of the present invention.

As used herein, the term tumor-associated antigen includes proteins or polypeptides that are preferentially expressed on the surface of a tumor cell. The expression preferentially expressed, as used herein, means that the antigen is expressed on the tumor cell at a level that is at least 10% greater (e.g., 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 150, 200, 400% or more) than the level of antigen expression on non-tumor cells. In certain embodiments, the target molecule is an antigen that is preferably expressed on the surface of a tumor cell selected from the group consisting of a kidney tumor cell, a colon tumor cell, a breast tumor cell, an ovarian tumor cell, a skin tumor cell, a lung tumor cell, prostate tumor cells, pancreatic tumor cells, glioblastoma cells, head and neck tumor cells and melanoma cells. Non-limiting examples of specific tumor-associated antigens include, for example, AFP, ALK, BAGE proteins, β-catenin, brc-abl, BRCA1, BORIS, CA9, carbonic anhydrase IX, caspase-8, CD40, CDK4, CEA, CTLA4, cyclin B1 , CYP1B1, EGFR, EGFRvIII, ErbB2/Her2,

- 12 046398

ErbB3, ErbB4, ETV6-AML, EphA2, Fra-1, FOLR1, GAGE proteins (e.g. GAGE-1, -2), GD2, GD3, GloboH, glypican-3, GM3, gp100, Her2, HLA/B-raf , HLA/k-ras, HLA/MAGE-A3, hTERT, LMP2, MAGE proteins (e.g. MAGE-1, -2, -3, -4, -6 and -12), MART-1, mesothelin, ML- IAP, Muc1, Muc16 (CA-125), MUM1, NA17, NY-BR1, NY-BR62, NY-BR85, NY-ESO1, OX40, p15, p53, PAP, PAX3, PAX5, PCTA-1, PLAC1, PRLR , PRAME, PSMA (FOLH1), RAGE proteins, Ras, RGS5, Rho, SART-1, SART-3, Steap-1, Steap-2, survivin, TAG-72, TGF-β, TMPRSS2, Tn, TRP-1 , TRP-2, tyrosinase and uroplakin-3.

The polyspecific antigen binding molecule in accordance with this aspect of the present invention may be conjugated to a drug, toxin, radioisotope or other substance that adversely affects cell viability. Alternatively, the drug or toxin may be a substance that does not directly kill the cell, but rather makes the cell more susceptible to killing by other foreign means. In still other embodiments involving tumor targeting, the multispecific antigen binding molecule of the present invention is not itself conjugated to a drug, toxin, or radioisotope, but is instead administered in combination with a second target (T) specific antigen binding molecule ( referred to herein as an accomplice molecule), wherein the accomplice molecule is conjugated to a drug, toxin, or radioisotope. In such embodiments, the polyspecific antigen binding molecule will preferably bind to an epitope on the target molecule (T) that is different from and/or does not overlap with the epitope recognized by the co-conspirator molecule (i.e., allowing simultaneous binding of the polyspecific antigen binding molecule and the molecule -accomplice with the target).

In a related embodiment, the present invention also includes antitumor combinations and therapeutic methods comprising (a) a toxin- or drug-conjugated antigen-binding molecule that specifically binds a tumor-associated antigen; and (b) a multispecific antigen-binding molecule comprising (i) a first binding domain that specifically binds an internalizing effector protein (eg, with low affinity), and (ii) a second binding domain that specifically binds a toxin- or drug-conjugated antigen-binding molecule.

In this embodiment, the multispecific antigen binding molecule functions to bind the toxin- or drug-conjugated antigen binding molecule to the internalizing effector protein, which thereby functions to physically link the tumor-associated antigen to the internalizing effector protein. Internalization of a toxin-labeled antibody to a tumor-associated antigen through its coupling to an internalizing effector protein would therefore result in targeted killing of tumor cells.

In accordance with certain embodiments of the tumor targeting aspects of the present invention, a multispecific antigen binding molecule (or co-conspirator antibody) may be conjugated to one or more cytotoxic drugs selected from the group consisting of calicheamicin, esperamycin, methotrexate, doxorubicin, melphalan, chlorambucil, ARA-C, vindesine, mitomycin C, cisplatin, etoposide, bleomycin, 5-fluorouracil, estramustine, vincristine, etoposide, doxorubicin, paclitaxel, larotaxel, tesetaxel, orataxel, docetaxel, dolastatin 10, auristatin E, auristat ina RNE and maytansine-based compounds (eg DM1, DM4, etc.). The polyspecific antigen binding molecule (or antibody co-conspirator) may also, or alternatively, be conjugated to a toxin such as diphtheria toxin, Pseudomonas aeruginosa exotoxin A, ricin A chain, abrin A chain, modessin A chain, alpha-sarcin, Aleurites proteins fordii, dianthin squirrels, Phytolaca americana squirrels, etc. The polyspecific antigen binding molecule (or antibody accomplice) may also, or alternatively, be conjugated to one or more radioisotopes selected from the group consisting of ²²⁵ Ac, ²¹¹ At, ²¹² Bi, ²¹³ Bi, ¹⁸⁶ Rh, ¹⁸⁸ Rh, ¹⁷⁷ Lu , ⁹⁰ Y, ¹³¹ I, ⁶⁷ Cu, ¹²⁵ I, ¹²³ I, ⁷⁷ Br, ¹⁵³ Sm, ¹⁶⁶ Ho, ⁶⁴ Cu, 121pb, ²²⁴ Ra and ²²³ Ra. Thus, this aspect of the present invention includes multispecific antigen-binding molecules that are antibody-drug conjugates (ADCs) or antibody-radioisotope conjugates (ARCs).

In the context of tumor killing applications, the D2 moiety may, under certain circumstances, bind with low affinity to the internalizing effector protein E. Thus, the multispecific antigen binding molecule will preferentially target tumor cells that express a tumor-associated antigen. As used herein, low affinity binding means that the binding affinity of the D2 component to the internalizing effector protein (E) is at least 10% weaker (e.g., 15% weaker, 25% weaker, 50% weaker, 75% weaker, 90% weaker, etc.) than the binding affinity of the D1 component to the target molecule (T). In certain embodiments it is possible to

- 13046398 low affinity binding means that D2-KOMnoHeHT interacts with internalizing effector protein (E) with a KD of greater than about 10 nM to about 1 μM, as measured in a surface plasmon resonance assay at about 25°C.

Simultaneous binding of a polyspecific antigen-binding molecule to an internalizing effector protein and a tumor-associated antigen will result in preferential internalization of the polyspecific antigen-binding molecule into tumor cells. If, for example, a multispecific antigen-binding molecule is conjugated to a drug, toxin, or radioisotope (or if a multispecific antigen-binding molecule is administered in combination with a co-conjugated antibody that is conjugated to a drug, toxin, or radioisotope), targeted internalization of the tumor-associated antigen into the tumor cell for due to its connection with a polyspecific antigen-binding molecule, it will lead to extremely specific destruction of tumor cells.

Pharmaceutical compositions and methods of administration.

The present invention includes pharmaceutical compositions containing a multispecific antigen binding molecule. The pharmaceutical compositions of the present invention can be formulated with suitable carriers, excipients and other means that provide improved transport, delivery, tolerability and the like.

The present invention also includes methods for inactivating or reducing the activity of a target molecule (T). The methods of the present invention involve contacting a target molecule with a multispecific antigen binding molecule as described herein. In certain embodiments, the methods of this aspect of the present invention include administering a pharmaceutical composition comprising a multispecific antigen binding molecule to a patient for whom it is necessary and/or beneficial to inactivate, attenuate, or otherwise reduce the extracellular concentration of the target molecule.

Various delivery systems are known in the art and can be used to administer the pharmaceutical compositions of the present invention to a patient. Routes of administration that may be used in the context of the present invention include, but are not limited to, intradermal, intramuscular, intraperitoneal, intravenous, subcutaneous, intranasal, epidural, and oral routes. The pharmaceutical compositions of the present invention can be administered by any convenient method, for example, by infusion or bolus injection, absorption through epithelial or mucocutaneous integuments (for example, oral mucosa, rectal and intestinal mucosa, etc.), and they can be administered together with other biologically active agents. Administration may be systemic or local. For example, the pharmaceutical composition of the present invention can be delivered subcutaneously or intravenously using a standard needle and syringe. Additionally, for subcutaneous delivery, a pen syringe can be used to administer the pharmaceutical composition of the present invention to a patient.

Examples

The following examples are set forth to provide those skilled in the art with complete disclosure and description of how to carry out and use the methods and preparation and use of the compositions of the present invention, and are not intended to limit the scope of what the inventors of the present invention consider their invention to be. Efforts have been made to ensure accuracy with respect to the numbers used (eg quantity, temperature, etc.), however, some experimental errors and deviations must be taken into account. Unless otherwise noted, parts are parts by weight, molecular weight is average molecular weight, temperature is in degrees Celsius, and pressure is at or near atmospheric.

Example 1 Use of a polyspecific antigen binding molecule to induce degradation of a cell surface receptor through binding to an internalizing effector protein.

In an initial proof-of-concept experiment, a multispecific antigen-binding molecule was created that is capable of binding (a) an internalizing effector molecule; and (b) the target molecule, which is a cell surface receptor.

In this example, the internalizing effector protein is Kremen-2 (Krm2) and the target cell surface receptor molecule is Fc receptor (FcyRI (Fc-gamma-R1)).

Kremen molecules (Krm1 and Krm2) are cell surface proteins known to mediate WNT signaling by controlling the internalization and degradation of LRP5 and LRP6 molecules involved in the WNT signaling pathway. LRP5/6 is internalized through the soluble interacting protein DKK1. In particular, DKK1 binds Kremen to LRP5/6 on the cell surface, and through this association, Kremen internalization controls the internalization and degradation of LRP5 and LRP6 (see Li et al., PLoS One, 5(6):e11014).

- 14 046398

The present inventors attempted to utilize the Kremen-binding properties of DKK1 and the internalization properties of Kremen to induce internalization of FcyRI. To promote Kremen-mediated internalization/degradation of FcyRI, a polyspecific antigen-binding molecule consisting of DKK1 fused to a mouse Fc (DKK1-mFc, containing the amino acid sequence of SEQ ID NO: 1) was constructed. As explained elsewhere herein, a multispecific antigen binding molecule is defined as a molecule comprising a first antigen binding domain (D1) that specifically binds a target molecule and a second antigen binding domain (D2) that specifically binds an internalizing effector protein. In this proof of concept example, the first antigen binding domain is an mFc component that specifically binds the target molecule FcyRI, and the second antigen binding domain is a DKK1 component that specifically binds the internalizing effector protein Kremen.

First, an experiment was performed to determine whether DKK1-mFc could be taken up by cells through endocytosis in a Kremen-dependent manner. Two cell lines were used for this experiment: Cell-1, a HEK293 cell line engineered to express FcyRI but not Kremen-2, and Cell-2, a HEK293 cell line engineered to express both FcyRI and Kremen-2. Conditioned medium containing DKKl-mFc at a dilution of 1:10 was added to the appropriate cell lines and incubated at 37°C for 90 min. After 90 min of incubation, cells were stained with A1exa-488-labeled anti-mouse IgG antibody to detect the DKK1-mFc molecule. Using fluorescence microscopy, it was found that DKK1-mFc was hardly localized to Cell-1 (without Kremen); however, significant amounts of DKK1-mFc were detected in Cell-2, which expresses Kremen-2. Thus, these results indicate that the multispecific antigen-binding molecule DKK1-mFc can be internalized into cells in a Kremen-dependent manner.

A time course experiment was then performed to determine whether DKK1-mFc could induce FcyRI degradation in a Kremen-dependent manner. A summary of the experimental protocol is as follows: Cell-1 (expressing FcyRI only) and Cell-2 (expressing Kremen-2 and FcyRI) were treated with 2 mg/ml sulfo-Ni^-biotin for 15 min on ice to label all expressed on cell surface proteins. Cells were then washed and resuspended in 400 μl of medium and divided into four 100 μl aliquots, which were treated with DKK1-mFc for various periods of time (0, 15, 30, and 60 min) at 37°C. After incubation with DKK1mFc, cells were pelleted and treated with protease inhibitors. Cell lysates for different incubation time points were subjected to FcyRI immunoprecipitation. For FcyRI immunoprecipitation, mouse anti-FcyRI antibody was added to cell lysates and incubated for 1 h at 4°C. Protein G beads were then added and the mixture was incubated for 1 h at 4°C. The beads were then washed and the proteins were eluted and subjected to SDS-PAGE. Proteins were transferred to the membrane and probed with HRP-labeled streptavidin to determine the relative amounts of remaining surface-presented FcyRI protein in each sample. The results are shown in Fig. 2.

As illustrated in FIG. 2, the amount of surface exposed FcyRI protein in Cell-1 samples (expressing FcyRI but not Kremen-2) remained relatively constant regardless of the length of time the cells were exposed to DKK1-mFc. In contrast, the amount of surface-presented FcyRI protein in Cell-2 samples (expressing both Kremen-2 and FcyRI) decreased significantly with increasing incubation time with DKK1-mFc. Thus, this experiment demonstrates that DKK1-mFc induces degradation of cell surface expressed FcyRI in a Kremen-2-dependent manner.

Taken together, the above results indicate that a multispecific antigen-binding molecule that simultaneously binds a cell surface target molecule (FcyRI) and an internalizing effector protein (Kremen-2) can induce degradation of the target molecule in an effector protein-dependent manner.

Example 2 IL-4R activity is attenuated using a multispecific antigen binding molecule with specificity for IL-4R and CD63.

In an additional series of proof-of-concept experiments, a polyspecific antigen-binding molecule was constructed that is capable of simultaneously binding a cell surface expressed target molecule (ie, IL-4R) and a cell surface expressed internalizing effector protein (ie, CD63). The purpose of these experiments was to determine whether the activity of IL-4R on a cell could be attenuated to a greater extent by physical binding of IL-4R to an effector molecule that is internalized and targeted for degradation in the lysosome (in this case CD63). In other words, this example was designed to test whether the normal internalization and degradation of CD63 could be used to stimulate the internalization and redirection of IL-4R for degradation within the cell.

First, a multispecific antigen-binding molecule was designed that is capable of binding both IL-4R and CD63. In particular, streptavidin-conjugated antibody to IL-4R and

- 15 046398 biotinylated anti-CD63 antibody was combined in a 1:1 ratio to form an anti-IL-4K:anti-CD63 conjugate (ie, a multispecific antigen-binding molecule that specifically binds IL-4R and CD63). The anti-IL-4R antibody used in this example is a complete human mAb raised against the extracellular domain of IL-4R. (The anti-IL-4R antibody contained a heavy chain variable region having SEQ ID NO: 3 and a light chain variable region having SEQ ID NO: 4). The anti-CD63 antibody used in this example is mouse anti-human CD63 mAb, clone MEM-259, obtained from Biolegend (San Diego, CA), catalog no. 312002.

Two control constructs were also generated: control-1=streptavidin-conjugated anti-IL-4R antibody combined 1:1 with biotinylated mouse anti-IgG1-knpa control antibody; and control-2=streptavidin-conjugated anti-IL-4R antibody combined 1:1 with non-biotinylated anti-CD63 antibody. The anti-IL-4R antibody used in the experimental and control constructs for this example is an antibody that is known to specifically bind IL-4R and only partially block IL-4-mediated signaling.

The experimental cell line used in this example is the HEK293 cell line containing a STAT6-luciferase reporter construct and additional STAT6 (HEK293/STAT6-luc cells). The cells used in this experiment express both IL-4R and CD63 on their surface. When treated with IL-4 in the absence of any inhibitors, this cell line produces a dose-dependent, detectable chemiluminescent signal that reflects the extent of IL-4-mediated signal transduction.

In the initial experiment, experimental IL-4R/CD63 binding polyspecific molecule or control constructs were added to HEK293/STAT6-luc cells such that the final concentration of anti-IL-4R antibody in the medium was 12.5 nM. The reporter signal was measured at increasing concentrations of IL-4 in the presence and absence of experimental and control constructs (Fig. 3). As can be seen in FIG. 3, a polyspecific IL-4R/CD63 binding molecule (ab-conjugate) inhibited IL-4-mediated signaling to a significantly greater extent than either control construct.

To confirm that the effect observed in FIG. 3, is CD63 dependent, the same experiment described above was performed, except that CD63 expression in the reporter cell line was significantly reduced using siRNA directed against CD63. With significantly reduced CD63 expression, increased inhibitory activity of the IL-4R/CD63 multispecific molecule was no longer observed (Fig. 4). This result suggests that the ability of the IL-4R/CD63-binding polyspecific molecule to attenuate IL-4-mediated signaling is due to simultaneous binding of the polyspecific molecule to IL-4R and CD63 and subsequent internalization and degradation of the entire antibody-IL-4R- complex. CD63.

Similar experiments were then performed in which the IL-4R/CD63 polyspecific binding molecule or control constructs were incubated with the HEK293/STAT6-luc reporter cell line for various periods of time before adding IL-4. In the first series of such experiments, the molecules were incubated with the reporter cell line for 0 h (ie, added simultaneously with IL-4), 2 h, or overnight before adding 50 pM IL-4. Luciferase activity was measured 6 h after the addition of IL-4. The results are shown in Fig. 5, top panel (untransfected). In another set of experiments, a similar protocol was performed except that the experimental or control molecules were incubated with the reporter cell line for 15, 30 min, 1, or 2 h before adding 50 pM IL-4. The results are shown in Fig. 6.

The results presented in Fig. 5 and 6 show that the multispecific IL-4R/CD63 binding molecule is able to inhibit IL-4-mediated signaling and that this inhibitory effect increases with longer incubation times. As in the original series of experiments, using siRNA for CD63 it was confirmed that the inhibitory effect of the polyspecific IL-4R/CD63 binding molecule was dependent on CD63 expression (Fig. 5, bottom panel (siRNA for CD63)).

Thus, this example provides further proof of concept of inhibiting the activity of a target molecule by using a multispecific antigen-binding molecule that is capable of simultaneously binding both the target molecule (in this case IL-4R) and the internalizing effector protein (in this case CD63), thereby thereby causing internalization and redirection for degradation of the target molecule within the cell. In other words, simultaneous binding of IL-4R and CD63 by an exemplary polyspecific antigen binding molecule attenuated IL-4R activity to a significantly greater extent (ie, >10%) than binding of IL-4R by control constructs alone.

Example 3: IL-4RxCD63 Bispecific Antibody Attenuates IL-4R Activity in a CD63-Dependent Manner.

The experiments of Example 2 herein show that the multispecific IL-4R/CD63 binding molecule inhibits IL-4-mediated CD63-dependent signaling.

- 16 046398 zom. In these experiments, the multispecific antigen-binding molecule consisted of two separate monoclonal antibodies (anti-IL-4R and anti-CD63) that were linked via a biotinstreptavidin linkage. In order to confirm that the results observed for this multispecific antigen binding molecule used for proof of concept are generalizable to other formats of multispecific antigen binding molecules, a true bispecific antibody was constructed.

Standard bispecific antibody technology was used to construct a bispecific antibody consisting of a first arm specific for IL-4R and a second arm specific for CD63. The IL-4R-specific arm contained an IL-4R-binding heavy chain paired with a CD63-specific light chain. The CD63-specific light chain was paired with the [HI-specific heavy chain solely for the purpose of convenience in obtaining the construct; however, pairing of the IL-4R-binding heavy chain with the CD63-binding light chain retained full specificity for IL-4R and did not exhibit CD63 binding. The CD63-specific arm contained a CD63-binding heavy chain paired with a CD63-binding light chain (the same light chain as in the IL-4R-binding arm). IL-4R binding heavy chain (containing SEQ ID NO: 3) was prepared from the complete anti-IL-4R antibody used in Example 2; and CD63-binding heavy and light chains were obtained from an anti-CD63 antibody designated H5C6 obtained from the Developmental Research Hybridoma Bank (University of Iowa, Department of Biology, Iowa City, Iowa). As with the complete anti-IL-4R antibody used in Example 2, the IL-4R binding component of the bispecific antibody used in this Example alone exhibited only moderate IL-4R blocking activity.

To evaluate the blocking activity of the IL-4RxCD63 bispecific antibody, an IL-4 luciferase assay was performed. Briefly, serial dilutions of IL-4RxCD63 bispecific antibody or control molecules were added to HEK293/STAT6-luc reporter cells (see Example 2). Under normal conditions, these cells produce a detectable luciferase signal when treated with IL-4. For this experiment, 10 pM IL-4 was added to the cells and luciferase activity was quantified for each dilution of antibody used. Control samples used in this analysis were as follows:

(a) a mimic bispecific antibody that binds IL-4R with one arm and has a nonfunctional CD63-binding arm (i.e., contains one IL-4R-binding heavy chain and one CD63-binding heavy chain, both paired with a light chain that binds to IL-4R);

(b) monospecific antibody to IL-4R; and (c) buffer (PBS) only (no antibody).

The results are shown in Fig. 7.

As shown in FIG. 7, for the control samples used, luciferase activity remained relatively high even at the highest antibody concentrations, whereas for the bispecific antibody, luciferase activity decreased significantly as antibody concentrations increased. These results confirm that simultaneous binding of IL-4R and CD63 by a bispecific antibody provides significant inhibition of IL-4R activity.

Example 4: Internalization of SOST using a multispecific antigen-binding molecule that simultaneously binds SOST and CD63.

In this example, the ability of multispecific antigen binding molecules to stimulate the internalization of a soluble SOST target molecule (sclerostin) was assessed. For these experiments, the target molecule was a fusion protein consisting of the human SOST protein tagged with pHrodo™ (Life Technologies, Carlsbad, CA) and a myc tag. pHrodo™ is a pH-sensitive dye that fluoresces substantially at neutral pH and fluoresces strongly in acidic environments such as the endosome. Therefore, the fluorescent signal can be used as an indicator of cellular internalization of the SOST-based fusion protein. The polyspecific antigen-binding molecules for these experiments were bispecific antibodies with binding specificities for both CD63 (internalizing effector protein) and SOST-based fusion protein (soluble target molecule), as described in more detail below.

The experiments were carried out as follows. Briefly, HEK293 cells were seeded at a concentration of 10,000 cells/well in poly-E-lysine-coated 96-well plates (Greiner Bio-One, Monroe, NC). After the cells had settled overnight, the medium was replaced with medium containing antibody (5 μg/ml as described below), pHrodo™-myc-labeled SOST (5 μg/ml), heparin (10 μg/ml) and Hoechst 33342 The cells were then incubated for 3 hours on ice or 3 hours at 37°C. All cells were washed twice before imaging in PBS, and the number of fluorescence spots per cell, as well as the corresponding fluorescence intensity, was counted to establish the extent of cellular internalization of pHrodo™-Tyc-labeled SOST in the presence of various antibody constructs.

- 17 046398

The antibodies used in this example were as follows:

(1) monospecific CD63 antibody (clone H5C6, Developmental Research Hybridoma Bank, University of Iowa, Department of Biology, Iowa City, Iowa);

(2) anti-myc antibody (clone 9E10, Schiweck et al., 1997, FEBS Lett., 414(1):33-38);

(3) an anti-SOST antibody (an antibody having the variable regions of the heavy and light chains of the antibody designated Ab-B in US Pat. No. 7,592,429);

(4) a bispecific antibody to CD63xmyc (ie, a multispecific antigen-binding molecule containing a CD63-binding arm derived from the H5C6 antibody and a myc-binding arm derived from 9E10);

(5) CD63xSOST bispecific antibody No. 1 (ie, a multispecific antigen-binding molecule comprising a CD63-binding arm derived from antibody H5C6 and an SOST-binding arm derived from Ab-B); and (6) CD63xSOST bispecific antibody No. 2 (i.e., a multispecific antigen-binding molecule comprising a CD63-binding arm derived from the H5C6 antibody and a SOST-binding arm derived from the antibody designated Ab-20 in U.S. Pat. No. 7592429).

The bispecific antibodies used in these experiments were collected using a methodology called projections into valleys (see, for example, Ridgway et al., 1996, Protein Eng., 9(7):617-621).

The results of the internalization experiments are shown in Fig. 8. In FIG. Figure 8 shows the number of spots (labeled vesicles) per cell under the various treatment conditions tested. Taken together, the results of these experiments demonstrate that bispecific constructs that simultaneously bind CD63 and SOST (either directly or via a myc tag) provided the greatest degree of SOST internalization, as reflected in fluorescence intensity and number of fluorescence spots per cell over time at 37°C. WITH. Thus, the multispecific antigen binding molecules used in this example are capable of effectively directing the internalization of a soluble target molecule.

Example 5 Changes in bone mineral density in mice treated with a multispecific antigen binding molecule that binds CD63 and SOST.

The CD63xSOST multispecific antigen binding molecule as described in Example 4 was then tested for its ability to increase bone mineral density in mice. Five groups of mice (approximately 6 mice per group) were used in these experiments. The treatment groups were as follows:

(I) untreated mice as negative control;

(ii) mice treated with a blocking monospecific antibody to SOST, which is known to itself increase bone mineral density (positive control);

(iii) mice treated with a bispecific antibody that specifically binds CD63 and SOST but does not itself inhibit SOST activity or only slightly inhibits SOST activity by itself;

(IV) mice treated with a parent CD63 antibody (ie, a monospecific antibody containing the same CD63-binding antigen-binding domain as the bispecific antibody); and (V) mice treated with a parent SOST antibody (i.e., a monospecific antibody containing the same SOST-binding antigen-binding domain as the bispecific antibody).

The amount of antibody administered to mice in each group was approximately 10-25 mg/kg.

It was expected that mice in Group III (treated with the bispecific anti-SOSTxCD63 antibody) would have an increase in bone mineral density that was at least comparable to that observed in mice in Group II (treated with a known blocking anti-SOST antibody), even though the bispecific antibody component to SOST does not inhibit SOST activity by itself (as evidenced by the mice in Group V, which were not expected to exhibit an increase in bone mineral density). The increase in bone mineral density that was expected in Group III mice is thought to be due to CD63-mediated internalization of SOST, as observed in the cell experiments of Example 4 above.

Example 6 Cellular internalization of lipopolysaccharide (LPS) is mediated by a multispecific antigen-binding molecule that simultaneously binds LPS and CD63.

This example illustrates the use of a multispecific antigen binding molecule of the present invention to control the internalization of a non-proteinaceous target molecule, namely lipopolysaccharide (LPS). LPS is a component of the outer membrane of gram-negative bacteria and is known to promote septic shock. Antibodies to LPS have been investigated as possible treatments for sepsis. The experiments in the present example were designed to evaluate the ability of a multispecific antigen-binding molecule to stimulate LPS internalization.

The multispecific antigen binding molecule used in this example was a bispecific antibody with one arm directed to LPS (target) and the other arm directed to CD63 (internalizing effector protein). The LPS binding arm was prepared

- 18 046398 from an antibody known as WN1 222-5 (DiPadova et al., 1993, Infection and Immunity, 61(9):3863-3872; Muller-Loennies et al., 2003, J. Biol. Chem., 278 (28):25618-25627; Gomery et al., 2012, Proc. Natl. Acad. Sci., USA, 109(51):20877-20882; US Patent No. 5858728). The CD63 binding arm was derived from the H5C6 antibody (see Example 4). Bispecific antibody to LPSxCD63 (ie, a multispecific antigen-binding molecule) was collected using a methodology called projections into valleys (see, for example, Ridgway et al., 1996, Protein Eng., 9(7):617-621).

Two types of LPS were used in these experiments: E. coli LPS and Salmonella minnesota LPS. Both versions were prepared as fluorescently labeled molecules (ALEXA-FLUOR®-488-labeled LPS, Life Technologies, Carlsbad, CA).

The experiments were carried out as follows. HEK293 cells were seeded in 96-well imaging plates coated with PDL. After resting overnight, the medium was replaced with fresh medium. Fluorescently labeled LPS (derived from either E. coli or S. minnesota) was added to standard medium. LPSxCD63 bispecific antibody or control half-antibodies paired with a mock Fc were then added to the samples. After various periods of incubation at 37°C (1 and 3 h) or on ice (3 h), cells from LPS-treated samples were treated as follows: washed - quenched with antibody to ALEXA-FLUOR®-488 - washed and fixed. Antibody to ALEXA-FLUOR®-488 quenches fluorescence from non-internalized (i.e., surface-bound) fluorophore. Thus, any fluorescence observed in samples treated with quenching antibody is due to internalized LPS. The fluorescence level from each sample was measured at different time points.

In fig. Figure 9 shows the results of these experiments in terms of the number of labeled vesicles per cell. As shown in FIG. 9, only cells treated with the bispecific antibody to CD63xLPS showed a significant number of labeled vesicles, which increased over time. For cells treated with labeled LPS and control antibodies, no significant number of fluorescent vesicles were detected, indicating that LPS was not internalized under these treatment conditions.

Thus, this example demonstrates that the LPSxCD63 bispecific antibody mediates the internalization of LPS into cells in a manner that requires simultaneous binding of LPS and CD63. Accordingly, these results provide a rationale for the use of the multispecific antigen binding molecules of the present invention to promote the cellular internalization of target molecules, such as LPS, for the treatment of diseases and disorders such as sepsis.

Example 7 Cellular internalization and degradation of Her2 mediated by a multispecific antigen binding molecule that simultaneously binds Her2 and PRLR.

This example includes three separate experiments that illustrate the use of the multispecific antigen binding molecule of the present invention to control the internalization and degradation of a tumor-associated antigen (Her2). Removal of Her2 from the surface of tumor cells would be a desirable approach for the treatment of certain cancers characterized by high expression of Her2 (eg, breast cancer, gastric cancer, gastroesophageal cancers, etc.).

The polyspecific antigen binding molecules used in the experiments of this example are bispecific antibodies with one arm directed to Her2 (target) and the other arm directed to PRLR (internalizing effector protein).

Experiment 1.

As a baseline experiment, the ability of a bispecific anti-Her2xPRLR antibody (herein referred to as bsAb1 anti-Her2xPRLR) to mediate degradation of endogenous Her2 in T47D breast cancer cells was assessed. T47D breast cancer cells express HER2 at low levels, and are generally not responsive to anti-HER2 therapies (see Horwitz et al., Variant T47D human breast cancer cells with high progesterone-receptor levels despite estrogen and antiestrogen resistance, Cell, 1982 Mar, 28(3):633-42).

The Her2 binding arm was derived from an antibody known as trastuzumab. The PRLR binding arm was derived from a complete anti-human PRLR antibody. Bispecific antibody to Her2xPRLR (ie, a multispecific antigen-binding molecule) was collected using a methodology called projections into valleys (see, for example, Ridgway et al., 1996, Protein Eng., 9(7):617-621). Other constructs used in this experiment were trastuzumab (a monospecific antibody), an anti-PRLR Ab (a monospecific antibody called H1H6958N2, set forth in US Patent Application Publication No. 2015/0056221, the disclosure of which is incorporated herein by reference in its entirety), and monospecific control antibody (directed against an irrelevant non-human antigen).

For this experiment, T47D cells expressing endogenous levels of Her2 receptors and PRLR were grown in RPMI (Irvine Scientific, 9160) supplemented with 10% FBS (ATCC, 30-2020), 10 mM Hepes, 1 mM sodium pyruvate, and 10 μg/ml insulin . Cycloheximide (CHX) at 50 μg/ml was used to stop protein synthesis. T47D cells were treated with either cycloheximide or trastuzumab

- 19 046398 (CHX/trastuzumab), cycloheximide and bsAbl to Her2xPRLR (CHX/bsAb1), cycloheximide and Ab to PRLR (CHX/PRLR Ab), cycloheximide and control Ab (CHX/control Ab), or cycloheximide, Ab to PRLR and trastuzumab for 0, 1, 2 or 4 hours.

Cells were lysed on ice in RIPA buffer (100 mM Tris-HCl, 300 mM NaCl, 2% NP-40, 1% sodium deoxycholate, and 0.2% SDS) (Boston BioProducts BP-116X) supplemented with protease and phosphatase inhibitors (Thermo Fisher, 1861280) and then sonicated (Qsonica Model Q55, three pulses). Sonicated lysates were diluted 2x with SDS sample buffer 1:1, heated at 95°C for 10 min, and then centrifuged at 13,000 rpm for 10 min at room temperature. Supernatants were separated on Novex 4–20% Tris-glycine gels and blotted onto PVDF membranes.

Anti-Her2 (Dako, A0485) at 1:300 or anti-PRLR (Life Technologies, 35-9200) at 1:250 was used for primary membrane labeling, followed by HRP-conjugated secondary antibody at 1:5000 and chemiluminescence detection using ECL (Amersham, RPN2106). Quantification for Western blot analyzes was performed by calculating net band intensity using CareStream software (Kodak). The results are shown in table. 1. Values are expressed in relative units compared to background.

As shown in table. 1, following inhibition of protein synthesis with cycloheximide, PRLR was shown to undergo rapid and complete degradation with all treatments used; the process was not disrupted by any of the antibodies tested. In contrast, in the presence of cycloheximide, Her2 was degraded in cells treated with bsAb1 to Her2xPRLR, but not with a mixture of trastuzumab and anti-PRLR Ab, individual anti-PRLR Ab, trastuzumab, or control antibodies. This result suggests that bsAb1 to Her2xPRLR serves to couple Her2 to PRLR at the cell surface, followed by degradation of the Her2-bsAb1-PRLR complex.

Table 1

Treatment Time (hours) Neg2 PRLR Cycloheximide+trastuzumab 0 5.87 6, 67 1 2.51 4.28 2 5.38 2.89 4 5.66 0.18 Cycloheximide+bsAbl to Her2xPRLR 0 3.99 6.48 1 5.79 3.68 2 2.36 1.62 4 0.70 0.04 Cycloheximide+Ab to PRLR 0 5.36 5.75 1 4.48 1.24 2 4.24 0.69 4 6.32 0 Cycloheximide+control antibody 0 3.36 4.66 1 3.09 3.10 2 4.98 3.30 4 4.74 0.23 Cycloheximide+trastuzumab+Ab to PRLR 0 5.37 6.12 1 4.22 2.16 2 4.48 0.44 4 5.22 0

Experiment 2.

The second experiment examined the dose-dependent effects of multispecific antigen-binding proteins binding to Her2xPRLR on Her2 degradation in vitro, as well as the potential impact on PRLR levels. Two different bispecific antibodies to Her2xPRLR (bsAb1 to Her2xPRLR and bsAb2 to Her2xPRLR) were used in this experiment. Both bispecific antibodies were constructed using the ridge-to-hole methodology. The Her2 binding arm was identical for both bispecific antibodies and was derived from an antibody known as trastuzumab. The PRLR binding arms for bsAbl to Her2xPRLR and bsAb2 to Her2xPRLR were derived from two different fully human anti-PRLR antibodies.

- 20 046398

For this experiment, T47D cells expressing endogenous levels of Her2 receptors and PRLR were grown in RPMI (Irvine Scientific, 9160) supplemented with 10% FBS (ATCC, 30-2020), 10 mM Hepes, 1 mM sodium pyruvate, and 10 μg/ml insulin . T47D cells were treated for 6 h with either bsAb1 or bsAb2 at 10, 3, 1, 0.3, 0.03, or 0 μg/ml.

Cells were lysed on ice in RIPA buffer (100 mM Tris-HCl, 300 mM NaCl, 2% NP-40, 1% sodium deoxycholate, and 0.2% SDS) (Boston BioProducts BP-116X) supplemented with protease and phosphatase inhibitors (Thermo Fisher, 1861280) and then sonicated (Qsonica Model Q55, three pulses). Sonicated lysates were diluted 2x with SDS sample buffer 1:1, heated at 95°C for 10 min, and then centrifuged at 13,000 rpm for 10 min at room temperature. Supernatants were separated on Novex 4–20% Tris-glycine gels and blotted onto PVDF membranes.

Anti-Her2 (Dako, A0485) at 1:300, anti-PRLR (Life Technologies, 35-9200) at 1:250, or anti-beta-actin (Genetex, GTX100313) at 1:10000 were used to initially label the membranes, then followed by HRP-conjugated secondary antibody at 1:5000 and chemiluminescence detection using ECL (Amersham, RPN2106).

Quantification for Western blot analyzes was performed by calculating net band intensity using CareStream software (Kodak). Normalization to actin was used as a loading control as follows: the sample with the highest net intensity for actin was used as the normalization control. The net intensity for each sample's actin was divided by the normalization control value to obtain a relative value for the sample. The net intensity values for each target protein (Her2 or PRLR) were divided by the calculated relative actin value for the sample to obtain a normalized Her2 or PRLR value (arbitrary units). The results are shown in table. 2.

As shown in table. 2, both bsAb1 to Her2xPRLR and bsAb2 to Her2xPRLR induced Her2 degradation in a dose-dependent manner in T47D cells. In contrast, PRLR levels were not affected by bispecific antibody treatment. This result is consistent with PRLR undergoing constitutive surface turnover on the surface of these cells.

table 2

Treatment Concentration (µg/ml) Neg2 PRLR bsAbl to Her2xPRLR 0 6, 47 2.49 0.03 3.44 0.95 oh 1 1.46 1, 10 0.3 1.47 1.33 10 1.10 1, 63 thirty 1.27 1, 64 10.0 0.98 1.30 bsAb2 to Her2xPRLR 0 6, 16 3, 48 0.03 4.37 2.71 oh 1 3.12 3, 64 0.3 2.43 3, 42 10 1.43 2.85 thirty 1.55 2.52 10.0 1.59 2.70

Experiment 3.

The above experiments suggest that degradation of Her2 by a Her2xPRLR bispecific antibody requires the bispecific antibody to simultaneously bind PRLR (internalizing effector protein) and Her2 (target protein). To confirm this principle, a third experiment was performed to evaluate the effect of blocking the PRLR- or Her2-binding arms of bsAb1 to Her2xPRLR on Her2 internalization activity.

For this experiment, T47D cells expressing endogenous levels of Her2 receptors and PRLR were grown in RPMI (Irvine Scientific, 9160) supplemented with 10% FBS (ATCC, 30-2020), 10 mM Hepes, 1 mM sodium pyruvate, and 10 μg/ml insulin . bsAb1 to Her2xPRLR (10 μg/ml) was left untreated (i.e., both arms free) or incubated with either a soluble PRLR protein construct (PRLR.mmh to block PRLR-binding arm binding) or a soluble Her2 protein construct (Her2. mmh to block the Her2-binding arm) at 37°C for 1 h (1:2 molar ratio) before adding to T47D cells for 0, 2, 4, or 6 h.

- 21 046398

At the end of the incubation period, cells were lysed on ice in RIPA buffer (100 mM Tris-HCl, 300 mM NaCl, 2% NP-40, 1% sodium deoxycholate, and 0.2% SDS) (Boston BioProducts BP-116X) supplemented with protease inhibitors and phosphatases (Thermo Fisher, 1861280) and then sonicated (Qsonica Model Q55, three pulses). Sonicated lysates were diluted 2x with SDS sample buffer 1:1 and heated at 95C for 10 min and then centrifuged at 13,000 rpm for 10 min at room temperature. Supernatants were separated on Novex 4–20% Tris-glycine gels and blotted onto PVDF membranes.

Anti-Her2 (Dako, A0485) at 1:300 or anti-beta-actin (Genetex, GTX100313) at 1:10,000 was used for primary membrane labeling, followed by HRP-conjugated secondary antibody at 1:5,000 and chemiluminescence detection by ECL (Amersham, RPN2106).

Quantification for Western blot analyzes was performed by calculating net band intensity using CareStream software (Kodak). Normalization to actin was used as a loading control as follows: the sample with the highest net intensity for actin was used as the normalization control. The net intensity for each sample's actin was divided by the normalization control value to obtain a relative value for the sample. The net intensity for the Her2 band was divided by the calculated relative actin value for the sample to obtain a normalized Her2 value (arbitrary units). The results are shown in table. 3.

As shown in table. 3, bsAb1 to Her2xPRLR-mediated degradation of Her2 was completely prevented in T47D cells by blocking either the Her2 or PRLR arms of bsAb1 to Her2xPRLR, indicating that degradation of Her2 occurred through its physical binding to PRLR, which was mediated by simultaneous binding of Her2 and PRLR by polyspecific antigen binding protein bsAbl Her2xPRLR.

Example 8: Enhanced activity of trastuzumab emtansine mediated by a multispecific antigen binding molecule that simultaneously binds Her2 and PRLR.

This example illustrates the use of the multispecific antigen binding molecules of the present invention to enhance the activity of an antibody-drug conjugate (ADC). More specifically, this example demonstrates the use of a multispecific antigen binding molecule comprising a first binding domain directed to an internalizing effector protein and a second binding domain directed to a tumor-associated target antigen, in combination with a second antigen binding molecule specific to the tumor-associated target antigen. (also referred to herein as an accomplice molecule), wherein the accomplice molecule is an ADC. In the experiments reported herein below, the internalizing effector protein is prolactin receptor (PRLR), the tumor-associated target antigen is Her2, and the co-conspirator molecule is a Her2-specific ADC (i.e., trastuzumab emtansine (T-DM1 )).

Trastuzumab is a recombinant humanized monoclonal antibody that binds to the extracellular domain of Her2. Trastuzumab and its corresponding ADC, trastuzumab emtansine or T-DM1, have been used successfully in patients with strong Her2 overexpression assessed by immunohistochemistry (IHC 3+). However, there is currently no effective therapy for patients with Her2 IHC2+ and Her2 IHC1+ tumors.

Table 3

Treatment Time (hours) Her2 bsAbl to Her2xPRLR 0 2.84 2 2.39 4 1.31 6 0.42 bsAbl to Her2xPRLR+PRLR.mmh 0 3.06 2 1, 94 4 2.65 6 2.30 bsAbl to Her2xPRLR+Her2.mmh 0 2.16 2 2.30 4 2.56 6 1.47

In this example, the ability of four different bispecific antibodies to Her2xPRLR (referred to herein as bsAb36 to Her2xPRLR, bsAb37 to Her2xPRLR, bsAb42

- 22 046398 to Her2xPRLR and bsAb45 to Her2xPRLR) provide an increase in the cytotoxic activity of T-DM1 on Her2-expressing cells. The bispecific antibodies used in this example were constructed from four different Her2-binding arms and two different PRLR-binding arms, as summarized in Table 1. 4.

Table 4

Bispecific antibody Shoulder connecting to Neg2 Shoulder connecting to PRLR bsalzb to Her2xPRLR AB-1 to Her2 Ab-1 to PRLR bsAL37 to Her2xPRLR Rd-2 to Neg2 Ab-1 to PRLR bsAL42 to Her2xPRLR AB-3 to Her2 Rd-2 to PRLR bsAb45 to Her2xPRLR AB-4 to Her2 Rb-2 to PRLR

Standard methods were used to construct bispecific antibodies to Her2xPRLR (ie, multispecific antigen-binding molecules). Control experiments were also performed using a control ADC (i.e., an antibody-drug conjugate containing an antibody directed against an irrelevant non-human protein conjugated to DM1) or an anti-PRLR-DM1 ADC (an antibody-drug conjugate referred to as H1H6958N2 in US 2015/0056221, the disclosure of which is incorporated herein by reference in its entirety).

To evaluate the ability of bispecific antibodies to Her2xPRLR to enhance the cytotoxic effect of T-DM1, T47D cells expressing endogenous levels of PRLR and overexpressing Her2 at intermediate levels (T47D/Her2) were first grown in RPMI (Irvine Scientific, 9160) supplemented with 10% FBS (ATCC, 30-2020), 10 mM Hepes, 1 mM sodium pyruvate and 10 μg/ml insulin. Cells were then seeded into 96-well plates at a concentration of 3000 cells/well. The next day, cells were left untreated or treated with a range of concentrations of either T-DM1, control ADC, anti-PRLR-DMi, or a combination of T-DM1 with 10 μg/ml bsAb42 to Her2xPRLR, bsAb36 to Her2xPRLR, bsAb37 to Her2xPRLR, or bsAb45 to Her2xPRLR for 3 days in triplicate.

Cell viability was assessed as follows. 3 days after treatment, cells were fixed with 0.25% PFA, 0.1% saponin, 2 μg/ml Hoechst for 20 min, whole-well images were acquired on an ImageXpress ^MICRO automated microscope at 10x and analyzed using the Cell Proliferation ^HT MetzxPress™ module Module (core counting) . The results are summarized in table. 5.

Table 5

Treatment IC50 (nM) T-DM1 30.0 Control ADC 200.0 T-DMl+bsAB42 to Her2xPRLR 2.0 Control ADC+bsAB42 to Her2xPRLR 100.0 ADC to PRLR 0.9 T-DMl+bsAB37 to Her2xPRLR 2.0 T-DMl+bsAB36 to Her2xPRLR 2.5 T-DMl+bsAB45 to Her2xPRLR 1.0

As shown in table. 5, the cytotoxicity activity of T-DM1 was significantly enhanced in the presence of the Her2xPRLR bispecific antibodies of the present invention. Thus, this example demonstrates the ability of the multispecific antigen binding proteins of the present invention to enhance the activity and effectiveness of immunoconjugate molecules directed to target proteins that are not typically rapidly internalized by cells.

T-DM1 is known to be active against tumors that express HER2 at high levels, whereas those cells that express HER2 at low and intermediate levels remain resistant to T-DM1 therapy. In contrast, anti-PRLR-ADC is able to induce robust killing of breast tumor cells expressing low levels of PRLR. The difference in killing efficiency between anti-PRLR-ADC and T-DM1 is likely explained by differences in the rate of internalization and lysosome translocation of their respective targets. The intracellular trafficking of PRLR and HER2 was compared.

In these cells expressing low levels of surface PRLR and HER2, PRLR was observed to undergo rapid constitutive lysosomal trafficking and degradation, which

- 23 046398 is mostly independent of the prolactin ligand. However, HER2 was not subject to rapid constitutive lysosomal trafficking and degradation. For example, in T47D cells, which express PRLR and HER2 at low levels, 80% of PRLR was internalized within 60 min, whereas HER was not internalized (Fig. 11). In these experiments, T47D cells were incubated on ice with either CF™594-labeled anti-PRLR primary antibody or CF™594-labeled anti-HER2 primary antibody. The internalization process was initiated by adding prewarmed (37°C) medium to the cells. At the indicated time points, cells were fixed with 4% paraformaldehyde to stop internalization and stained with a secondary Alexa Fluor® 488-conjugated goat anti-human Fab antibody to identify non-internalized antibodies remaining on the cell surface (yellow). Co-localization was quantified on a pixel-by-pixel basis for all sections of the confocal series. The results are shown in Fig. eleven.

Similarly, PRLR, but not HER2, was found to be rapidly internalized into the lysosomal compartment of T47D cells. CF™594-labeled anti-PRLR or CF™594-labeled anti-HER2 antibodies were added to T47D cells prelabeled with 3 kDa fluorescein dextrans. The antibody-receptor complexes were allowed to internalize at 37°C for the indicated times (X-axis in Fig. 12), then fixed with 4% paraformaldehyde. The co-localization of internalized antibodies for dextran-labeled lysosomes was determined. The results are shown in Fig. 12 and are essentially consistent with the internalization data in FIG. eleven.

Sequence motifs determining constitutive PRLR renewal were mapped to a 21-amino acid region in the cytoplasmic domain of the receptor; PRLR resumption ensured effective ADC-mediated cell cycle arrest and cell killing. In fig. 13 depicts various PRLR and HER2 constructs used in the following experiments, including (1) full-length PRLR (PRLR FL);

(2) full-length HER2 (HER2 FL);

(3) the extracellular (Ecto) domain of PRLR fused to the transmembrane (TM) and cytoplasmic (Cyto) portions of HER2 (PRLRectoHER2cyto™);

(4) the extracellular (Ecto) domain of HER2 fused to the transmembrane (TM) and cytoplasmic (Cyto) portions of PRLR (HER2ectoPRLRcyto™);

(5) PRLR with amino acids 301-622 removed (as measured by the remaining full-length protein with signal peptide) (PRLR Cyto42); and (6) PRLR with amino acids 280–622 (as measured by the remaining full-length protein with signal peptide) deleted (PRLR Cyto21).

To map PRLR internalization effector sequences to one of its domains (extracellular, transmembrane, cytoplasmic), HEK293 cells were grown in 24-well optical bottom plates and transiently transfected with a mammalian expression vector encoding PRLR, HER2, HER2ectoPRLRcyto™, or PRLRectoHER2cyto™, for 24 hours. Transfected cells were then incubated on ice in the presence of CF™594-labeled primary antibody and then warmed to 37°C and incubated for an additional 1 hour. FIG. 14 shows stained cells at 4°C, in the case where internalization of the constructs is not expected), and after 37°C, in the case when those constructs that can be quickly internalized will be internalized. Panel A depicts the internalized PRLR FL construct at 37°C. Panel B shows the HER2 FL construct, which was essentially not internalized within 1 h at 37°C. Panel C shows the PRLRecto-HER2cyto/TM construct, which was also largely not internalized within 1 h at 37°C. Panel D shows the internalized HER2ectoPRLRcyto/TM construct at 37°C. These results suggest that the cytoplasmic domain of PRLR provides the internalization effector sequence.

The internalization effector sequence was further mapped to PRLR amino acid residues from about 280 to about 300 (e.g., residues 280-300 of SEQ ID NO: 11, residues DAHLLEKGKSEELLSALGCQD (SEQ ID NO: 12)) (also known as PRLR Cyto21-42). Experiments with labeled antibodies from ice-at-37°C were performed on HEK293 cells as described above. In fig. Figure 15 depicts the cellular uptake of those constructs that contain the complete cytoplasmic domain of PRLR (PRLR FL) (panels A and B), membrane-proximal 42 amino acids (PRLR Cyto42) (panels C and D), and membrane-proximal 21 amino acids (PRLR Cyto21) ( panels D and E). While the PRLR Cyto42 construct was rapidly internalized, such internalization was absent for the PRLR Cyto21 construct. This demonstrates that the region of the PRLR cytoplasmic domain between amino acids 21 and 42 proximal to the plasma membrane (Cyto21–42) contains the information required for PRLR internalization and degradation.

To determine whether internalized PRLR constructs were degraded (i.e., targeted to the lysosome), transfected HEK293 cells were treated with cycloheximide (CHX) at various time points after heating (0, 2, 4, 6 h) to stop protein production. As demonstrated by Western blotting of total cell lysates (Fig. 16), PRLR FL and PRLR Cyto42 rapidly

- 24 046398 were degraded, while almost no degradation of PRLR Cyto21 was observed.

The effectiveness of the transmembrane and cytoplasmic domain of PRLR as an internalization effector for anti-cell proliferation drug delivery in the form of an antibody-drug conjugate (ADC) was tested in transfected HEK293 cells. HEK293 cells were engineered to express PRLR, HER2, or PRLRectoHER2cyto/TM in a tetracycline-controlled manner (eg, using the Lenti-X™ Tet-On® system, Clontech, Mountain View, CA). Transfected cells were induced for 24 h with 0.01 or 0.003 or 0.7 μg/ml doxycycline to achieve different levels of receptor expression, which were determined by flow cytometry. Either PRLR conjugated to mertansine (also known as DM1, also known as emtansine) (1 nM) (panel A, Fig. 17), or HER2DM1 (1 nM) (panel B, Fig. 17), was added to the corresponding transfected cells which were incubated for an additional 24 h. Cell cycle analysis was performed using an antibody to phosphohistone H3 (Ser10) to detect cells in early mitosis. The results are shown in Fig. 17, which demonstrated that the transmembrane/cytoplasmic domain of PRLR effectively mediates drug delivery into cells.

Alanine mutagenesis of the cytoplasmic domain, in particular the Cyto21-24 domain, was performed to map residues required for internalization. Site-directed mutagenesis of two dileucines contained in the Cyto21-42 sequence inhibited the resumption of Cyto42 PRLR on transfected HEK293 cells. T-REx™ HEK293 cells (i.e., cells stably expressing Tet-On, ThermoFisher Scientific, Waltham, MA) were engineered to express PRLR Cyto42, PRLR Cyto21, PRLR Cyto42 L283A L284A, PRLR Cyto42 L292A L293A, or PRLR Cyto42 L283A L284A L292A L293A (also called 4LA) in a tetracycline-controlled manner (using, for example, the Jump-In™ Cell Engineering platform, ThermoFisher Scientific, Waltham, MA), was induced with doxycycline (0.7 μg/ml) for 24 h and treated with CHX (50 µg/ml) for 0 or 4 hours. The results are shown in the Western blot in Fig. 18, which demonstrates reduced protein degradation for constructs lacking Cyto21-42 or dileucine repeats. The results support the conclusion that these residues are important for lysosomal targeting.

Additional experiments demonstrated that PRLR turnover, mediated by its cytoplasmic domain, is the driving force targeting HER2 for lysosomal degradation when both receptors are coupled by bispecific antibodies to HER2(T)xPRLR. (HERZfr) is the arm of trastuzumab.) First, the binding kinetic parameters of the bispecific antibody and its parent monospecific antibodies were assessed. The bispecific antibody showed similar binding kinetics to the extracellular domains of both HER2 and PRLR as its parent antibodies (Table 6).

To demonstrate that the cytoplasmic domain of PRLR is required for HER2(T)xPRLR bsAb-mediated degradation of endogenous HER2, T-REx™ HEK293 cells were engineered to express either full-length PRLR or a truncated PRLR lacking the entire cytoplasmic domain (lower panel). Proteins were expressed in a tetracycline-controlled manner using the Cell-Platform Jump-In™ (ThermoFisher Scientific, Waltham, MA). Cells were induced for 24 h with 0.7 μg/ml doxycycline, followed by incubation with CHX in combination with the indicated antibodies (Table 6) for various periods of time, lysed and processed for Western blotting.

Table 6

Antibody Protein Kinetic binding parameters k _a (M ^-1 c ^_1 ) kd(c-1) _KD (M) Tl/2 (4.) Antibody to HER2(T) hHER2.mmh 2.54xl0 ⁵ 1.01x10-3 3, 98x10-9 eleven Anti-PRLR antibody hPRLR. mmh 1.20x10° 4.15x10-3 3.45x10-9 3 Bispecific to HER2(T) x PRLR hHER2.mmh 1.70xl0 ⁶ 2.86xl0- ⁴ 1.69x10-9 40 hPRLR.mmh 5.51X10 ⁶ 3.36x10-4 6, lOxlO- ¹⁰ 34

In fig. 19 shows Western blot results of lysates from cells expressing the full-length PRLR construct. Notably, the anti-HER2(T)xPRLR bispecific antibody efficiently targets HER2 to the lysosome, whereas the anti-HER2(T) antibody does not, demonstrating the effectiveness of PRLR as an internalization effector for HER2 expressed at low levels. In fig. 20 shows Western blot results of lysates from cells expressing the cytoplasmic truncation PRLR construct. However, no degradation of PRLR or HER2 was noted with any treatment, demonstrating the requirement for the cytoplasmic domain of PRLR to target this receptor and its molecular cargo to the lysosome.

Bispecific antibody to HER2xPRLR induced the translocation of HER2 into the lysosome, increased

- 25 046398

HEB2-AOC.'-induced cell cycle arrest and promoted the killing of T47D cells. To determine lysosome targeting, T47D cells were treated with either nonbinding control antibodies or a bispecific anti-HER2xPRLR antibody and subjected to an ice-to-37°C endocytosis assay. Lysosomes and HER2 were stained and the percentage of lysosomal marker pixels occupied by HER2 pixels was determined. The results are shown in Fig. 21, which demonstrated that more than 50–60% of HER2 was targeted to the lysosome by the HER2xPRLR bispecific antibody within 60 min of heating compared with less than 25% for controls.

To determine that HER2xPRLR bispecific antibody enhances HER2-DM1-mediated cell cycle arrest, HER2-expressing T47D cells were treated with 1 nM, 10 nM, or 30 nM PRLR-ADC (A), HER2-ADC plus HER2xPRLR bispecific antibody (B) , HER2-ADC alone (C), nonbinding control ADC (D), or no treatment (E). The percentage of cells in early mitosis was determined by staining with an antibody to phospho-histone (Ser10), an indicator of cell cycle arrest. The results are shown in Fig. 22, which demonstrated that HER2-ADC plus a bispecific antibody to HER2xPRLR (B) was more effective in cell arrest than HER2-ADC alone (C).

To determine that the HER2xPRLR bispecific antibody enhances HER2DM1-mediated cell killing, HER2-expressing T47D cells were treated with drug as shown in FIG. 23: PRLR-ADC (A), HER2-ADC plus HER2xPRLR bispecific antibody (B), HER2-ADC alone (C), non-binding ADC control (D), or non-binding ADC control plus anti-HER2xPRLR bispecific antibody (E). In fig. 23 shows the results and demonstrates that the HER2xPRLR bispecific antibody significantly reduces the IC50 of HER2-ADC.

Example 9. High rate of target renewal and degradation using hMHC-1 as an internalizing molecule.

Proteins with high turnover rates were used to degrade soluble and transmembrane targets.

Multispecific antigen-binding proteins (e.g., bispecific antibodies) are designed to bind the target molecule to degradation proteins (i) that are known to regenerate rapidly, (ii) that have high levels of target-mediated clearance of bivalent antibodies, and (iii) that are known to translocate into or out of the lysosome. To facilitate development and manufacture, bispecific antibodies were created containing a conventional light chain, a heavy chain directed to the destructive protein, and a heavy chain directed to the target. One of the heavy chains contained the H95R modification.

In this example, the selected disruptive molecule is major histocompatibility complex I, MHC-1, more specifically class I isoform B (also known as HLA-B). Stable binding anti-HLA monoclonal antibodies are rapidly cleared, which is a hallmark of a disruptive molecule. As an example of a soluble target molecule, the FelD1 allergen was selected as the target molecule. FelD1 is a tetrameric glycoprotein containing two heterodimers linked by disulfide bonds. Two forms of soluble FelD1, a FelD1-myc-myc-his fusion protein (FelD1-mmh) and a FelD1-Fc fusion protein, were constructed and tested.

In one set of experiments, FelD1-mmh was tagged with Alexa 488 to facilitate tracking of target internalization. A bispecific antibody containing an HLA-B-specific arm (binding a destructive molecule) and a FelD1-specific arm (binding a target) (a-HLAB27-a-FelD1) was used as a polyspecific antigen-binding protein. C1Rneo B lymphoblastoid cells expressing HLA-B were incubated with 10 μg/ml FelD1-mmh-Alexa 488 and 10 μg/ml a-HLAB27-aFelD1. Cells were incubated overnight to allow time for internalization of the target FelD1 protein. Cells were then stained with an anti-Alexa488 antibody (Alexa-Fluor-488-Antibody-Polyclonal/A-11094, Thermo Fisher Scientific, Waltham, MA), which quenches the fluorescence of Alexa488. Antibody to Alexa488 will not quench a labeled target that has been internalized, so an internalized target can be distinguished from a target that is bound on the cell surface. Here, fluorescence was quantified using FACS.

In fig. 10 shows the average fluorescence (FACS arbitrary units) of surface-bound and internalized target for both MHC1-negative cells that serve as controls and MHC1-positive cells. The bispecific antibody used was a-HLAB27-a-FelD1. The original bivalent antibody a-HLAB27-a-HLAB27 and a nonspecific isotype IgG were used as controls. For cells expressing MHC1 (a destructive molecule), cells contacting a-HLAB27-a-FelD1 showed an approximately fourfold increase in internalized target molecule relative to surface-associated target molecule. Almost no effect was found for controls. The results are shown in Fig. 10 and in table. 7 below.

α-HLAB27-α-FelD1-mediated clearance of FelD1-Fc from the blood serum of mice was assessed.

- 26 046398 which expresses the human HLA-B allele. Bispecific antibody a-HLAB27-a-FelD1 and controls (PBS, bivalent monospecific antibody to FelD1 and bivalent monospecific antibody to HLAB27) were administered to mice expressing the human HLA-B allele by subcutaneous injection (10 mg/kg). The next day, 1.6 mg/kg FelD1-Fc was administered by tail vein injection. (Antibody:target ratio of 3:1 was used). Serum samples were obtained from tail vein blood collected at 15 min, 6 h, 1, 2, 3, 4, 6, and 8 days. FelD1 levels were detected and quantified by Western blotting using anti-FelD1 antibodies. Treatment with bispecific a-HLAB27-a-FelD1 demonstrated rapid clearance of FelD1 with t1/2 <30 h, similar to the clearance rate of anti-HLAB27 antibody in the absence of FelD1 (i.e. t1/2 33 h) and more than twice the clearance rate a-HLAB27-a-FelD1 in the absence of FelD1 (i.e. t1/2 65 h). Administration of anti-FelD1 antibody did not affect MHC1-mediated clearance, but some moderate clearance was observed, which was attributed to Fc receptors. The results are shown in Fig. 11 and 12.

Table 7

Average Fluorescence Units (FelD1-mmh-Alexa288)

Antibody B-lymphoblastoid cells C1R peo MHC minus cells MHC plus cells On a surface Internalized On a surface Internalized OC-HLAB27 "α-FelDl < 500 < 500 ~825 ~3350 OC-HLAB27 "α-FelDl < 500 < 500 < 500 < 500 Isotype control < 500 < 500 < 500 < 500

Example 10 Target degradation using the APLP2/PCSK9 system.

To evaluate whether proprotein convertase subtilisin/kexin type 9 (PCSK9) and its cell surface partners (such as LDLR and APLP2; see DeVay et al., Characterization of proprotein convertase subtilisin/kexin type 9 (PCSK9) trafficking reveals a novel lysosomal targeting mechanism via amyloid precursor-like protein 2 (APLP2), 288(15), J. Biol. Chem., 10805-18 (Apr 12, 2013)) used as effectors for antibody-mediated target destruction, fusions were obtained antitarget PCSK9 proteins. The fusion protein can be considered as a model for a bispecific antibody that includes a PSCK9 or APLP2 binding arm instead of a PCSK9 fusion.

For these experiments, hemeuvelin (HJV) was used as a model target in part due to the fact that the in vivo potency readout, i.e. Measuring the increase in serum iron one week after the start of treatment is simple and reliable. HJV is a coreceptor for bone morphogenetic protein 6 (BMP6). Blocking HJV inhibits BMP6 signaling and reduces hepcidin levels, which in turn inhibits the iron transporter ferroporitin. Ultimately, blocking HJV increases serum iron (see Core et al., Hemojuvelin and bone morphogenetic protein (BMP) signaling in iron homeostasis, 5 Front. Pharmacol., 104(1-9), May 13, 2014 .)

Six antibody::PCSK9 fusion molecules were generated, each containing either a non-blocking anti-HJV antibody (α-HJV-n) or an anti-Myc antibody (oc-Myc) in the hIgG1 backbone fused to one of three different PCSK9 variants (e.g. SEQ ID NO: 5) using a 3x GGGS linker (SEQ ID NO: 6). The first variant is full-length PCSK9 without a signal sequence, but including a pro-domain (full-length (PCSK9FL); SEQ ID NO: 7). The second variant is only the C-terminal domain, including a defined internal linker sequence between the catalytic and C-terminal domains of PCSK9 (long C-terminal (PCSK9LC), SEQ ID NO: 8). The third variant is a short variant of the C-terminal domain that does not include this internal linker sequence (short C-terminal (PCSK9SC), EQ ID NO: 9). The C-terminal variants are expected to bind only to APLP2 and not to the LDL receptor.

Antibody::PCSK9 hybrids were expressed and secreted by CHO cells. Plasmids containing both the heavy chain of the PCSK9 fusion and the cognate light chain of the corresponding antibody were cotransfected into CHO-K1 cells in 10 or 15 cm dishes. The cells were then incubated for 4–5 days to allow production and secretion of the fusion antibody: :PCSK9, after which the supernatant was collected and sterilized by filtration through 0.2-micron filters. CHO cell supernatants containing the fusion proteins were subsequently tested for their ability to internalize soluble HJV protein in vitro.

Soluble HJV labeled with pHrodo fluorophore was prepared as follows. HJV ecto-domain

-

Claims (26)

(SEQ ID NO: 10), fused to the myc-myc-6xhis (mmh) tag using a GPG linker, was expressed in CHO supernatants and purified. The purified protein was then labeled with pHrodo® (Thermo Fisher Scientific, Waltham) using the chemical N-hydroxysulfosuccinimide (NHS). HepG2 cells were incubated with a solution containing 50% antibody::PCSK9-containing CHO supernatant, 50% HepG2 medium with 5% lipoprotein-poor serum, and 1 μg/ml pHrodo®-labeled hHJV-mmh. HepG2 cells are a convenient model because they express both LDLR and APLP2, key proteins known to internalize PCSK9. However, to evaluate PCSK9 fusion polyspecific antigen binding proteins, one skilled in the art will understand that any cell line (natural, induced, ectopically transformed, etc.) in which PCSK9 is internalized, including through binding partners other than LDLR and APLP2 can be used in this assay. Cells were seeded at 1.5 × 10 ⁵ cells per ml the day before analysis and incubated at 37°C. Internalization was monitored over a 24-hour period by imaging on the Imager ImageXpress® for simultaneous multiparameter analysis (Molecular Devices, Sunnyvale, CA), with the 24-hour time point showing the largest difference between constructs and controls, likely due to accumulation of pHrodo®-HJV-mmh in cells throughout the analysis. Fluorescence reflecting internalization was quantified using MetaXpress® software (Molecular Devices). Incubation with a-HJV-N::PCSK9FL and a-Myc::PCSK9FL increased the degree of internalization of pHrodo®-tagged HJV-mmh protein compared to the control without antibodies. Because pHrodo-tagged HJV-mmh contains two myc tags, a-Myc::PCSK9FL is expected to perform similarly to a-HJV-N::PCSK9FL in this assay. Interestingly, fusion proteins containing the C-terminal domains of PCSK9 (i.e., a-HJV-N::PCSK9LC and a-HJV-N::PCSK9SC) failed to internalize the tagged HJV protein, possibly due to reduced binding LDLR. The results are shown in Fig. 26. Example 11 In Vivo Target Degradation Using the APLP2/PCSK9 System. The original α-HJV-N bivalent monospecific antibody (which does not block BMP binding to HJV), α-HJV-B bivalent monospecific antibody (which blocks BMP binding to HJV), a-HJV-N::PCSK9FL fusion protein, and a fusion protein -Myc::PCSK9FL were tested for their ability to block HJV signaling in vivo. The molecules were administered to mice using hydrodynamic delivery (HDD). Serum samples were collected after a week. Because endogenous HJV does not have a myc-myc-his tag, the α-Myc::PCSK9FL fusion protein functions as a negative control in vivo. As expected, HDD of the bivalent multispecific antibody α-HJV-N (positive control) resulted in a significant increase in serum iron relative to negative controls. HDD a-HJV-N::PCSK9FL (test molecule) also resulted in a significant increase in serum iron, similar in magnitude to that of the blocking antibody, suggesting that this molecule is effective in internalizing and either sequestering or destroying HJV in vivo (Fig. 27). The scope of the present invention is not limited to the specific embodiments described herein. Indeed, various modifications of the present invention other than those described herein will be apparent to those skilled in the art from the foregoing description and accompanying figures. Such modifications are within the scope of the appended claims. CLAIM 1. A method for ensuring degradation in a lysosome of a tumor-associated antigen (TAA) expressed on the surface of a breast cancer cell, including the step of physically binding TAA to the prolactin receptor (PRLR) by contacting the breast cancer cell with a polyspecific antigen-binding protein, which contains (a) antigen-binding fragment of anti-TAA antibody; and (b) an antigen binding fragment of an anti-PRLR antibody, wherein the TAA is selected from the group consisting of AFP, ALK, β-catenin, CA9, CD40, CEA, EGFR, ErbB3, ErbB4, EphA2, FOLR1, Gd3, GloboH, MAGE protein, mesothelin , Muc1 and PRAME, where the antibody binds to TAA and PRLR on the same breast cancer cell, and where TAA is transported to the lysosome through its physical association with PRLR and degraded. 2. The method according to claim 1, wherein the polyspecific antigen-binding protein is a bispecific antibody. 3. The method according to claim 1, where the cell is in vivo. 4. The method of claim 1, wherein the PRLR expressed on the surface of the breast cancer cell undergoes constitutive surface renewal. 5. A method for destroying a breast cancer cell that expresses a TAA molecule and a destructive PRLR molecule on its surface, the method comprising the steps - 28 046398 (i) contacting a breast cancer cell with a bivalent monospecific anti-TAA antibody conjugated to a cytotoxic agent, and (ii) physically binding TAA to a disruptive PRLR molecule on the breast cancer cell from part (i) by contacting the cell with the antibody, which contains (a) an antigen-binding fragment of an anti-TAA antibody; and (b) an antigen binding fragment of an anti-PRLR antibody to simultaneously bind to TAA and PRLR on the same cell, wherein the TAA molecule and the cytotoxic agent, through a physical link between the TAA molecule and the disruptive PRLR molecule, are then transported to a lysosome in said breast cancer cell thereby allowing for degradation of TAA, and wherein the TAA is selected from the group consisting of AFP, ALK, β-catenin, CA9, CD40, CEA, EGFR, ErbB3, ErbB4, EphA2, FOLR1, GD3, GloboH, MAGE protein, mesothelin, Muc1 and PRAME. 6. The method of claim 5, wherein the breast cancer cell is a T47D cell. 7. The method according to claim 5, where the breast cancer cell is in vivo. 8. The method according to claim 5, wherein the cytotoxic activity of the cytotoxic agent is enhanced by the presence of the antibody. 9. The method of claim 5, wherein the antibody is a bispecific antibody. 10. The method according to claim 5, where the cytotoxic agent is a radioisotope, toxin or drug. 11. The method of claim 10, wherein the toxin is calicheamicin, auristatin or a maytansine-based cytotoxin. 12. The method of claim 5, wherein the PRLR degradative molecule expressed on the surface of the breast cancer cell undergoes constitutive surface renewal. 13. A method for destroying a breast cancer cell that expresses on its surface a TAA molecule and a destructive PRLR molecule, the method comprising the step of physically binding TAA to PRLR by contacting the cell with a conjugate of an antibody and a cytotoxic agent, which contains (a) an antigen-binding fragment of an antibody to TAA; (b) an antigen binding fragment of an anti-PRLR antibody; and (c) a cytotoxic agent to simultaneously bind to TAA and PRLR on the same cell, wherein the TAA molecule and the cytotoxic agent, through a physical link between the TAA molecule and the disruptive PRLR molecule, are then transported to a lysosome in said breast cancer cell, thereby allowing degradation of TAA, and wherein the TAA is selected from the group consisting of AFP, ALK, β-catenin, CA9, CD40, CEA, EGFR, ErbB3, ErbB4, EphA2, FOLR1, GD3, GloboH, MAGE protein, mesothelin, Muc1 and PRAME. 14. The method of claim 13, wherein the breast cancer cell is a T47D cell. 15. The method according to claim 13, where the breast cancer cell is in vivo. 16. The method according to claim 13, wherein the cytotoxic activity of the cytotoxic agent is enhanced by the presence of the antibody. 17. The method of claim 13, wherein the antibody is a bispecific antibody. 18. The method according to claim 13, where the cytotoxic agent is a radioisotope, toxin or drug. 19. The method of claim 18, wherein the toxin is calicheamicin, auristatin or a maytansine-based cytotoxin. 20. The method of claim 13, wherein the PRLR degradative molecule expressed on the surface of the breast cancer cell undergoes constitutive surface renewal. 21. A method of inhibiting the growth of a tumor or promoting the regression of a tumor that contains a breast cancer cell that expresses both a TAA molecule and a destructive PRLR molecule on its surface, the method comprising killing the breast cancer cell according to the method of claim 5. 22. The method according to claim 21, wherein the TAA is selected from the group consisting of ALK, EGFR, ErbB3 and ErbB4. 23. The method of claim 21, wherein the antibody that physically binds TAA to PRLR is a bispecific anti-TAAxPRLR antibody. 24. The method according to claim 21, where the cytotoxic agent is a maytansine-based cytotoxin. 25. A method of inhibiting the growth of a tumor or promoting the regression of a tumor that contains a breast cancer cell that expresses on its surface both a TAA molecule and a destructive PRLR molecule, the method comprising killing the breast cancer cell according to the method of claim 13. 26. The method of claim 25, wherein the TAA is selected from the group consisting of ALK, EGFR, ErbB3 and ErbB4. -